Authors: Atahan Karagoz
Abstract: Survival is traditionally modeled as a supervised learning task, reliant on curated outcome labels and fixed covariates. This work rejects that premise. It proposes that survival is not an externally annotated target but a geometric consequence: an emergent property of the curvature and flow inherent in biological state space. We develop a theory of Self-Organizing Survival Manifolds (SOSM), in which survival-relevant dynamics arise from low-curvature geodesic flows on latent manifolds shaped by internal biological constraints. A survival energy functional based on geodesic curvature minimization is introduced and shown to induce structures where prognosis aligns with geometric flow stability. We derive discrete and continuous formulations of the objective and prove theoretical results demonstrating the emergence and convergence of survival-aligned trajectories under biologically plausible conditions. The framework draws connections to thermodynamic efficiency, entropy flow, Ricci curvature, and optimal transport, grounding survival modeling in physical law. Health, disease, aging, and death are reframed as geometric phase transitions in the manifold's structure. This theory offers a universal, label-free foundation for modeling survival as a property of form, not annotation-bridging machine learning, biophysics, and the geometry of life itself.
Authors: Fatemeh Moradi, Mehran Tarif, Mohammadhossein Homaei
Abstract: Detecting fraud in modern supply chains is a growing challenge, driven by the complexity of global networks and the scarcity of labeled data. Traditional detection methods often struggle with class imbalance and limited supervision, reducing their effectiveness in real-world applications. This paper proposes a novel two-phase learning framework to address these challenges. In the first phase, the Isolation Forest algorithm performs unsupervised anomaly detection to identify potential fraud cases and reduce the volume of data requiring further analysis. In the second phase, a self-training Support Vector Machine (SVM) refines the predictions using both labeled and high-confidence pseudo-labeled samples, enabling robust semi-supervised learning. The proposed method is evaluated on the DataCo Smart Supply Chain Dataset, a comprehensive real-world supply chain dataset with fraud indicators. It achieves an F1-score of 0.817 while maintaining a false positive rate below 3.0%. These results demonstrate the effectiveness and efficiency of combining unsupervised pre-filtering with semi-supervised refinement for supply chain fraud detection under real-world constraints, though we acknowledge limitations regarding concept drift and the need for comparison with deep learning approaches.
Authors: Azmine Toushik Wasi
Abstract: Drug-drug interactions pose a significant challenge in clinical pharmacology, with severe class imbalance among interaction types limiting the effectiveness of predictive models. Common interactions dominate datasets, while rare but critical interactions remain underrepresented, leading to poor model performance on infrequent cases. Existing methods often treat DDI prediction as a binary problem, ignoring class-specific nuances and exacerbating bias toward frequent interactions. To address this, we propose a framework combining Generative Flow Networks (GFlowNet) with Variational Graph Autoencoders (VGAE) to generate synthetic samples for rare classes, improving model balance and generate effective and novel DDI pairs. Our approach enhances predictive performance across interaction types, ensuring better clinical reliability.
Authors: A. Quadir, M. Tanveer
Abstract: In recent years, graph neural networks (GNNs) have gained significant attention for node classification tasks on graph-structured data. However, traditional GNNs primarily focus on adjacency relationships between nodes, often overlooking the rich role-based characteristics that are crucial for learning more expressive node representations. Existing methods for capturing role-based features are largely unsupervised and fail to achieve optimal performance in downstream tasks. To address these limitations, we propose a novel hypergraph neural network with state space model (HGMN) that effectively integrates role-aware representations into GNNs and the state space model. HGMN utilizes hypergraph construction techniques to model higher-order relationships and combines role-based and adjacency-based representations through a learnable mamba transformer mechanism. By leveraging two distinct hypergraph construction methods-based on node degree and neighborhood levels, it strengthens the connections among nodes with similar roles, enhancing the model's representational power. Additionally, the inclusion of hypergraph convolution layers enables the model to capture complex dependencies within hypergraph structures. To mitigate the over-smoothing problem inherent in deep GNNs, we incorporate a residual network, ensuring improved stability and better feature propagation across layers. Extensive experiments conducted on one newly introduced dataset and four benchmark datasets demonstrate the superiority of HGMN. The model achieves significant performance improvements on node classification tasks compared to state-of-the-art GNN methods. These results highlight HGMN's ability to provide enriched node representations by effectively embedding role-based features alongside adjacency information, making it a versatile and powerful tool for a variety of graph-based learning applications.
Authors: Zian Zhai, Fan Li, Xingyu Tan, Xiaoyang Wang, Wenjie Zhang
Abstract: Vector Quantization (VQ) has recently emerged as a promising approach for learning discrete representations of graph-structured data. However, a fundamental challenge, i.e., codebook collapse, remains underexplored in the graph domain, significantly limiting the expressiveness and generalization of graph tokens.In this paper, we present the first empirical study showing that codebook collapse consistently occurs when applying VQ to graph data, even with mitigation strategies proposed in vision or language domains. To understand why graph VQ is particularly vulnerable to collapse, we provide a theoretical analysis and identify two key factors: early assignment imbalances caused by redundancy in graph features and structural patterns, and self-reinforcing optimization loops in deterministic VQ. To address these issues, we propose RGVQ, a novel framework that integrates graph topology and feature similarity as explicit regularization signals to enhance codebook utilization and promote token diversity. RGVQ introduces soft assignments via Gumbel-Softmax reparameterization, ensuring that all codewords receive gradient updates. In addition, RGVQ incorporates a structure-aware contrastive regularization to penalize the token co-assignments among similar node pairs. Extensive experiments demonstrate that RGVQ substantially improves codebook utilization and consistently boosts the performance of state-of-the-art graph VQ backbones across multiple downstream tasks, enabling more expressive and transferable graph token representations.
Authors: Xinglin Zhao, Yanwen Wang, Xiaobo Liu, Yanrong Hao, Rui Cao, Xin Wen
Abstract: Computer-aided diagnosis (CAD) systems play a crucial role in analyzing neuroimaging data for neurological and psychiatric disorders. However, small-sample studies suffer from low reproducibility, while large-scale datasets introduce confounding heterogeneity due to multiple disease subtypes being labeled under a single category. To address these challenges, we propose a novel federated learning framework tailored for neuroimaging CAD systems. Our approach includes a dynamic navigation module that routes samples to the most suitable local models based on latent subtype representations, and a meta-integration module that combines predictions from heterogeneous local models into a unified diagnostic output. We evaluated our framework using a comprehensive dataset comprising fMRI data from over 1300 MDD patients and 1100 healthy controls across multiple study cohorts. Experimental results demonstrate significant improvements in diagnostic accuracy and robustness compared to traditional methods. Specifically, our framework achieved an average accuracy of 74.06\% across all tested sites, showcasing its effectiveness in handling subtype heterogeneity and enhancing model generalizability. Ablation studies further confirmed the importance of both the dynamic navigation and meta-integration modules in improving performance. By addressing data heterogeneity and subtype confounding, our framework advances reliable and reproducible neuroimaging CAD systems, offering significant potential for personalized medicine and clinical decision-making in neurology and psychiatry.
Authors: Rachel K. Luu, Jingyu Deng, Mohammed Shahrudin Ibrahim, Nam-Joon Cho, Ming Dao, Subra Suresh, Markus J. Buehler
Abstract: Large language models (LLMs) have reshaped the research landscape by enabling new approaches to knowledge retrieval and creative ideation. Yet their application in discipline-specific experimental science, particularly in highly multi-disciplinary domains like materials science, remains limited. We present a first-of-its-kind framework that integrates generative AI with literature from hitherto-unconnected fields such as plant science, biomimetics, and materials engineering to extract insights and design experiments for materials. We focus on humidity-responsive systems such as pollen-based materials and Rhapis excelsa (broadleaf lady palm) leaves, which exhibit self-actuation and adaptive performance. Using a suite of AI tools, including a fine-tuned model (BioinspiredLLM), Retrieval-Augmented Generation (RAG), agentic systems, and a Hierarchical Sampling strategy, we extract structure-property relationships and translate them into new classes of bioinspired materials. Structured inference protocols generate and evaluate hundreds of hypotheses from a single query, surfacing novel and experimentally tractable ideas. We validate our approach through real-world implementation: LLM-generated procedures, materials designs, and mechanical predictions were tested in the laboratory, culminating in the fabrication of a novel pollen-based adhesive with tunable morphology and measured shear strength, establishing a foundation for future plant-derived adhesive design. This work demonstrates how AI-assisted ideation can drive real-world materials design and enable effective human-AI collaboration.
Authors: Kyle O'Brien, Stephen Casper, Quentin Anthony, Tomek Korbak, Robert Kirk, Xander Davies, Ishan Mishra, Geoffrey Irving, Yarin Gal, Stella Biderman
Abstract: Open-weight AI systems offer unique benefits, including enhanced transparency, open research, and decentralized access. However, they are vulnerable to tampering attacks which can efficiently elicit harmful behaviors by modifying weights or activations. Currently, there is not yet a robust science of open-weight model risk management. Existing safety fine-tuning methods and other post-training techniques have struggled to make LLMs resistant to more than a few dozen steps of adversarial fine-tuning. In this paper, we investigate whether filtering text about dual-use topics from training data can prevent unwanted capabilities and serve as a more tamper-resistant safeguard. We introduce a multi-stage pipeline for scalable data filtering and show that it offers a tractable and effective method for minimizing biothreat proxy knowledge in LLMs. We pretrain multiple 6.9B-parameter models from scratch and find that they exhibit substantial resistance to adversarial fine-tuning attacks on up to 10,000 steps and 300M tokens of biothreat-related text -- outperforming existing post-training baselines by over an order of magnitude -- with no observed degradation to unrelated capabilities. However, while filtered models lack internalized dangerous knowledge, we find that they can still leverage such information when it is provided in context (e.g., via search tool augmentation), demonstrating a need for a defense-in-depth approach. Overall, these findings help to establish pretraining data curation as a promising layer of defense for open-weight AI systems.
Authors: Fangjun Hu, Guangkuo Liu, Yifan Zhang, Xun Gao
Abstract: As a class of generative artificial intelligence frameworks inspired by statistical physics, diffusion models have shown extraordinary performance in synthesizing complicated data distributions through a denoising process gradually guided by score functions. Real-life data, like images, is often spatially structured in low-dimensional spaces. However, ordinary diffusion models ignore this local structure and learn spatially global score functions, which are often computationally expensive. In this work, we introduce a new perspective on the phases of data distributions, which provides insight into constructing local denoisers with reduced computational costs. We define two distributions as belonging to the same data distribution phase if they can be mutually connected via spatially local operations such as local denoisers. Then, we show that the reverse denoising process consists of an early trivial phase and a late data phase, sandwiching a rapid phase transition where local denoisers must fail. To diagnose such phase transitions, we prove an information-theoretic bound on the fidelity of local denoisers based on conditional mutual information, and conduct numerical experiments in a real-world dataset. This work suggests simpler and more efficient architectures of diffusion models: far from the phase transition point, we can use small local neural networks to compute the score function; global neural networks are only necessary around the narrow time interval of phase transitions. This result also opens up new directions for studying phases of data distributions, the broader science of generative artificial intelligence, and guiding the design of neural networks inspired by physics concepts.
Authors: Md Arafat Hossain, Xingfu Wu, Valerie Taylor, Ali Jannesari
Abstract: Over the past few years, the size of language models has grown exponentially, as has the computational cost to train these large models. This rapid growth has motivated researchers to develop new techniques aimed at enhancing the efficiency of the training process. Despite these advancements, optimally predicting the model size or allocating optimal resources remains a challenge. Several efforts have addressed the challenge by proposing different scaling laws, but almost all of them are architecture-specific (dense or sparse). In this work we revisit existing scaling laws and propose a generalized scaling law to provide a unified framework that is applicable to both dense and sparse large language models. We evaluate and compare our proposed scaling law with existing scaling laws to demonstrate its effectiveness.
Authors: Jeremiah Birrell, Reza Ebrahimi
Abstract: We introduce ANTIDOTE, a new class of objectives for learning under noisy labels which are defined in terms of a relaxation over an information-divergence neighborhood. Using convex duality, we provide a reformulation as an adversarial training method that has similar computational cost to training with standard cross-entropy loss. We show that our approach adaptively reduces the influence of the samples with noisy labels during learning, exhibiting a behavior that is analogous to forgetting those samples. ANTIDOTE is effective in practical environments where label noise is inherent in the training data or where an adversary can alter the training labels. Extensive empirical evaluations on different levels of symmetric, asymmetric, human annotation, and real-world label noise show that ANTIDOTE outperforms leading comparable losses in the field and enjoys a time complexity that is very close to that of the standard cross entropy loss.
Authors: Mosbah Aouad, Anirudh Choudhary, Awais Farooq, Steven Nevers, Lusine Demirkhanyan, Bhrandon Harris, Suguna Pappu, Christopher Gondi, Ravishankar Iyer
Abstract: Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest cancers, and early detection remains a major clinical challenge due to the absence of specific symptoms and reliable biomarkers. In this work, we propose a new multimodal approach that integrates longitudinal diagnosis code histories and routinely collected laboratory measurements from electronic health records to detect PDAC up to one year prior to clinical diagnosis. Our method combines neural controlled differential equations to model irregular lab time series, pretrained language models and recurrent networks to learn diagnosis code trajectory representations, and cross-attention mechanisms to capture interactions between the two modalities. We develop and evaluate our approach on a real-world dataset of nearly 4,700 patients and achieve significant improvements in AUC ranging from 6.5% to 15.5% over state-of-the-art methods. Furthermore, our model identifies diagnosis codes and laboratory panels associated with elevated PDAC risk, including both established and new biomarkers. Our code is available at https://github.com/MosbahAouad/EarlyPDAC-MML.
Authors: Yewon Byun, Shantanu Gupta, Zachary C. Lipton, Rachel Leah Childers, Bryan Wilder
Abstract: Predictions and generations from large language models are increasingly being explored as an aid to computational social science and human subject research in limited data regimes. While previous technical work has explored the potential to use model-predicted labels for unlabeled data in a principled manner, there is increasing interest in using large language models to generate entirely new synthetic samples (also termed as synthetic simulations), such as in responses to surveys. However, it is not immediately clear by what means practitioners can combine such data with real data and yet produce statistically valid conclusions upon them. In this work, we introduce a new estimator based on generalized method of moments, providing a hyperparameter-free solution with strong theoretical guarantees to address the challenge at hand. Surprisingly, we find that interactions between the moment residuals of synthetic data and those of real data can improve estimates of the target parameter. We empirically validate the finite-sample performance of our estimator across different regression tasks in computational social science applications, demonstrating large empirical gains.
Authors: Muyan Anna Li, Aditi Gautam
Abstract: As time series data become increasingly prevalent in domains such as manufacturing, IT, and infrastructure monitoring, anomaly detection must adapt to nonstationary environments where statistical properties shift over time. Traditional static thresholds are easily rendered obsolete by regime shifts, concept drift, or multi-scale changes. To address these challenges, we introduce and empirically evaluate two novel adaptive thresholding frameworks: Segmented Confidence Sequences (SCS) and Multi-Scale Adaptive Confidence Segments (MACS). Both leverage statistical online learning and segmentation principles for local, contextually sensitive adaptation, maintaining guarantees on false alarm rates even under evolving distributions. Our experiments across Wafer Manufacturing benchmark datasets show significant F1-score improvement compared to traditional percentile and rolling quantile approaches. This work demonstrates that robust, statistically principled adaptive thresholds enable reliable, interpretable, and timely detection of diverse real-world anomalies.
Authors: Jonas S Almeida, Daniel E Russ, Susana Vinga, Ines Duarte, Lee Mason, Praphulla Bhawsar, Aaron Ge, Arlindo Oliveira, Jeya Balaji Balasubramanian
Abstract: Motivation: With the advent of Language Models using Transformers, popularized by ChatGPT, there is a renewed interest in exploring encoding procedures that numerically represent symbolic sequences at multiple scales and embedding dimensions. The challenge that encoding addresses is the need for mechanisms that uniquely retain contextual information about the succession of individual symbols, which can then be modeled by nonlinear formulations such as neural networks. Context: Universal Sequence Maps(USM) are iterated functions that bijectively encode symbolic sequences onto embedded numerical spaces. USM is composed of two Chaos Game Representations (CGR), iterated forwardly and backwardly, that can be projected into the frequency domain (FCGR). The corresponding USM coordinates can be used to compute a Chebyshev distance metric as well as k-mer frequencies, without having to recompute the embedded numeric coordinates, and, paradoxically, allowing for non-integers values of k. Results: This report advances the bijective fractal encoding by Universal Sequence Maps (USM) by resolving seeding biases affecting the iterated process. The resolution had two results, the first expected, the second an intriguing outcome: 1) full reconciliation of numeric positioning with sequence identity; and 2) uncovering the nature of USM as an efficient numeric process converging towards a steady state sequence embedding solution. We illustrate these results for genomic sequences because of the convenience of a planar representation defined by an alphabet with only 4 tokens (the 4 nucleotides). Nevertheless, the application to alphabet of arbitrary cardinality was found to be straightforward.
Authors: Andrey Sidorenko, Paul Tiwald
Abstract: Synthetic data generation has become essential for securely sharing and analyzing sensitive data sets. Traditional anonymization techniques, however, often fail to adequately preserve privacy. We introduce the Tabular Auto-Regressive Generative Network (TabularARGN), a neural network architecture specifically designed for generating high-quality synthetic tabular data. Using a discretization-based auto-regressive approach, TabularARGN achieves high data fidelity while remaining computationally efficient. We evaluate TabularARGN against existing synthetic data generation methods, showing competitive results in statistical similarity, machine learning utility, and detection robustness. We further perform an in-depth privacy evaluation using systematic membership-inference attacks, highlighting the robustness and effective privacy-utility balance of our approach.
Authors: Fernando Martinez-Lopez, Tao Li, Yingdong Lu, Juntao Chen
Abstract: Reinforcement learning (RL) agents often struggle to generalize to new tasks and contexts without updating their parameters, mainly because their learned representations and policies are overfit to the specifics of their training environments. To boost agents' in-context RL (ICRL) ability, this work formulates ICRL as a two-agent emergent communication problem and introduces CORAL (Communicative Representation for Adaptive RL), a framework that learns a transferable communicative context by decoupling latent representation learning from control. In CORAL, an Information Agent (IA) is pre-trained as a world model on a diverse distribution of tasks. Its objective is not to maximize task reward, but to build a world model and distill its understanding into concise messages. The emergent communication protocol is shaped by a novel Causal Influence Loss, which measures the effect that the message has on the next action. During deployment, the previously trained IA serves as a fixed contextualizer for a new Control Agent (CA), which learns to solve tasks by interpreting the provided communicative context. Our experiments demonstrate that this approach enables the CA to achieve significant gains in sample efficiency and successfully perform zero-shot adaptation with the help of pre-trained IA in entirely unseen sparse-reward environments, validating the efficacy of learning a transferable communicative representation.
Authors: Yuan-Hung Chao, Chia-Hsun Lu, Chih-Ya Shen
Abstract: Graph Neural Networks (GNNs) have shown strong performance on graph-structured data, but their reliance on graph connectivity often limits scalability and efficiency. Kolmogorov-Arnold Networks (KANs), a recent architecture with learnable univariate functions, offer strong nonlinear expressiveness and efficient inference. In this work, we integrate KANs into three popular GNN architectures-GAT, SGC, and APPNP-resulting in three new models: KGAT, KSGC, and KAPPNP. We further adopt a multi-teacher knowledge amalgamation framework, where knowledge from multiple KAN-based GNNs is distilled into a graph-independent KAN student model. Experiments on benchmark datasets show that the proposed models improve node classification accuracy, and the knowledge amalgamation approach significantly boosts student model performance. Our findings highlight the potential of KANs for enhancing GNN expressiveness and for enabling efficient, graph-free inference.
Authors: Chia-Hsun Lu, Guan-Jhih Wu, Ya-Chi Ho, Chih-Ya Shen
Abstract: With the increasing importance of protecting intellectual property in machine learning, watermarking techniques have gained significant attention. As advanced models are increasingly deployed in domains such as social network analysis, the need for robust model protection becomes even more critical. While existing watermarking methods have demonstrated effectiveness for conventional deep neural networks, they often fail to adapt to the novel architecture, Kolmogorov-Arnold Networks (KAN), which feature learnable activation functions. KAN holds strong potential for modeling complex relationships in network-structured data. However, their unique design also introduces new challenges for watermarking. Therefore, we propose a novel watermarking method, Discrete Cosine Transform-based Activation Watermarking (DCT-AW), tailored for KAN. Leveraging the learnable activation functions of KAN, our method embeds watermarks by perturbing activation outputs using discrete cosine transform, ensuring compatibility with diverse tasks and achieving task independence. Experimental results demonstrate that DCT-AW has a small impact on model performance and provides superior robustness against various watermark removal attacks, including fine-tuning, pruning, and retraining after pruning.
Authors: Md. Akmol Masud, Md Abrar Jahin, Mahmud Hasan
Abstract: Federated Learning (FL) is a machine learning technique that often suffers from training instability due to the diverse nature of client data. Although utility-based client selection methods like Oort are used to converge by prioritizing high-loss clients, they frequently experience significant drops in accuracy during later stages of training. We propose a theoretical HeteRo-Select framework designed to maintain high performance and ensure long-term training stability. We provide a theoretical analysis showing that when client data is very different (high heterogeneity), choosing a smart subset of client participation can reduce communication more effectively compared to full participation. Our HeteRo-Select method uses a clear, step-by-step scoring system that considers client usefulness, fairness, update speed, and data variety. It also shows convergence guarantees under strong regularization. Our experimental results on the CIFAR-10 dataset under significant label skew ($\alpha=0.1$) support the theoretical findings. The HeteRo-Select method performs better than existing approaches in terms of peak accuracy, final accuracy, and training stability. Specifically, HeteRo-Select achieves a peak accuracy of $74.75\%$, a final accuracy of $72.76\%$, and a minimal stability drop of $1.99\%$. In contrast, Oort records a lower peak accuracy of $73.98\%$, a final accuracy of $71.25\%$, and a larger stability drop of $2.73\%$. The theoretical foundations and empirical performance in our study make HeteRo-Select a reliable solution for real-world heterogeneous FL problems.
Authors: Hager Radi Abdelwahed, M\'elisande Teng, Robin Zbinden, Laura Pollock, Hugo Larochelle, Devis Tuia, David Rolnick
Abstract: Species distribution models (SDMs) are widely used to predict species' geographic distributions, serving as critical tools for ecological research and conservation planning. Typically, SDMs relate species occurrences to environmental variables representing abiotic factors, such as temperature, precipitation, and soil properties. However, species distributions are also strongly influenced by biotic interactions with other species, which are often overlooked. While some methods partially address this limitation by incorporating biotic interactions, they often assume symmetrical pairwise relationships between species and require consistent co-occurrence data. In practice, species observations are sparse, and the availability of information about the presence or absence of other species varies significantly across locations. To address these challenges, we propose CISO, a deep learning-based method for species distribution modeling Conditioned on Incomplete Species Observations. CISO enables predictions to be conditioned on a flexible number of species observations alongside environmental variables, accommodating the variability and incompleteness of available biotic data. We demonstrate our approach using three datasets representing different species groups: sPlotOpen for plants, SatBird for birds, and a new dataset, SatButterfly, for butterflies. Our results show that including partial biotic information improves predictive performance on spatially separate test sets. When conditioned on a subset of species within the same dataset, CISO outperforms alternative methods in predicting the distribution of the remaining species. Furthermore, we show that combining observations from multiple datasets can improve performance. CISO is a promising ecological tool, capable of incorporating incomplete biotic information and identifying potential interactions between species from disparate taxa.
Authors: Connor Brown
Abstract: First-order methods underpin most large-scale learning algorithms, yet their classical convergence guarantees hinge on carefully scheduled step-sizes that depend on the total horizon $T$, which is rarely known in advance. The Schedule-Free (SF) method promises optimal performance with hyperparameters that are independent of $T$ by interpolating between Polyak--Ruppert averaging and momentum, but nonconvex analysis of SF has been limited or reliant on strong global assumptions. We introduce a robust Lyapunov framework that, under only $L$-smoothness and lower-boundedness, reduces SF analysis to a single-step descent inequality. This yields horizon-agnostic bounds in the nonconvex setting: $O(1/\log T)$ for constant step + PR averaging, $O(\log T/T)$ for a linearly growing step-size, and a continuum of $O(T^{-(1-\alpha)})$ rates for polynomial averaging. We complement these proofs with Performance Estimation Problem (PEP) experiments that numerically validate our rates and suggest that our $O(1/\log T)$ bound on the original nonconvex SF algorithm may tighten to $O(1/T)$. Our work extends SF's horizon-free guarantees to smooth nonconvex optimization and charts future directions for optimal nonconvex rates.
Authors: Xingke Yang, Liang Li, Sicong Li, Liwei Guan, Hao Wang, Xiaoqi Qi, Jiang Liu, Xin Fu, Miao Pan
Abstract: Collaboratively fine-tuning (FT) large language models (LLMs) over heterogeneous mobile devices fosters immense potential applications of personalized intelligence. However, such a vision faces critical system challenges. Conventional federated LLM FT approaches place prohibitive computational and memory burdens on mobile hardware, and their synchronous model aggregation protocols stall for slower devices. In this paper, we propose Fed MobiLLM, a novel design to facilitate efficient federated LLM FT across mobile devices with diverse computing/communication speeds and local model architectures. In particular, Fed MobiLLM implements a pioneering server-assisted federated side-tuning paradigm. Briefly, mobile devices perform lightweight forward propagation computations on local data using their frozen pre-scaled backbone LLMs, and then upload selected intermediate activations. The server trains a shared side-network independently, eliminating client-side backpropagation and enabling asynchronous updates. To bridge model heterogeneity across different devices, we introduce an adaptive layer-wise feature alignment method, which ensures consistent representations for collaboratively tuning a shared side network. Extensive experimental results demonstrate that Fed MobiLLM can maintain robust fine-tuning performance while achieving extremely low on-device memory, with at least 95.2% reduction in computation overhead, 93.2% reduction in communication costs and 5.1x faster convergence compared to existing methods, validating its efficacy for practical LLM adaptation over heterogeneous mobile devices.
Authors: Arman Dogru, R. Irem Bor-Yaliniz, Nimal Gamini Senarath
Abstract: Digital Twins (DTs) are transforming industries through advanced data processing and analysis, positioning the world of DTs, Digital World, as a cornerstone of nextgeneration technologies including embodied AI. As robotics and automated systems scale, efficient data-sharing frameworks and robust algorithms become critical. We explore the pivotal role of data handling in next-gen networks, focusing on dynamics between application and network providers (AP/NP) in DT ecosystems. We introduce PANAMA, a novel algorithm with Priority Asymmetry for Network Aware Multi-agent Reinforcement Learning (MARL) based multi-agent path finding (MAPF). By adopting a Centralized Training with Decentralized Execution (CTDE) framework and asynchronous actor-learner architectures, PANAMA accelerates training while enabling autonomous task execution by embodied AI. Our approach demonstrates superior pathfinding performance in accuracy, speed, and scalability compared to existing benchmarks. Through simulations, we highlight optimized data-sharing strategies for scalable, automated systems, ensuring resilience in complex, real-world environments. PANAMA bridges the gap between network-aware decision-making and robust multi-agent coordination, advancing the synergy between DTs, wireless networks, and AI-driven automation.
Authors: Amit Pandey
Abstract: We present Zero-Direction Probing (ZDP), a theory-only framework for detecting model drift from null directions of transformer activations without task labels or output evaluations. Under assumptions A1--A6, we prove: (i) the Variance--Leak Theorem, (ii) Fisher Null-Conservation, (iii) a Rank--Leak bound for low-rank updates, and (iv) a logarithmic-regret guarantee for online null-space trackers. We derive a Spectral Null-Leakage (SNL) metric with non-asymptotic tail bounds and a concentration inequality, yielding a-priori thresholds for drift under a Gaussian null model. These results show that monitoring right/left null spaces of layer activations and their Fisher geometry provides concrete, testable guarantees on representational change.
Authors: Noel Teku, Fengwei Tian, Payel Bhattacharjee, Souradip Chakraborty, Amrit Singh Bedi, Ravi Tandon
Abstract: Alignment is a key step in developing Large Language Models (LLMs) using human feedback to ensure adherence to human values and societal norms. Dependence on human feedback raises privacy concerns about how much a labeler's preferences may reveal about their personal values, beliefs, and personality traits. Existing approaches, such as Differentially Private SGD (DP-SGD), provide rigorous privacy guarantees by privatizing gradients during fine-tuning and alignment but can provide more privacy than necessary as human preferences are tied only to labels of (prompt, response) pairs and can degrade model utility. This work focuses on LLM alignment with preference-level privacy, which preserves the privacy of preference labels provided by humans. We propose PROPS (PROgressively Private Self-alignment), a multi-stage privacy preserving alignment framework where privately aligned models in previous stages can serve as labelers for supplementing training data in the subsequent stages of alignment. We present theoretical guarantees for PROPS as well as comprehensive validation using multiple models (Pythia and GPT) and datasets (AlpacaEval, Anthropic HH-RLHF, truthy-dpo-v0.1) to demonstrate the utility of PROPS over existing methods while still providing high privacy. For the same privacy budget, alignment via PROPS can achieve up to 3x higher win-rates compared to DP-SGD, and 2.5x higher win-rates compared to Randomized Response (RR) based alignment.
Authors: Junjing Zheng, Chengliang Song, Weidong Jiang, Xinyu Zhang
Abstract: High-dimensional data, particularly in the form of high-order tensors, presents a major challenge in self-supervised learning. While MLP-based autoencoders (AE) are commonly employed, their dependence on flattening operations exacerbates the curse of dimensionality, leading to excessively large model sizes, high computational overhead, and challenging optimization for deep structural feature capture. Although existing tensor networks alleviate computational burdens through tensor decomposition techniques, most exhibit limited capability in learning non-linear relationships. To overcome these limitations, we introduce the Mode-Aware Non-linear Tucker Autoencoder (MA-NTAE). MA-NTAE generalized classical Tucker decomposition to a non-linear framework and employs a Pick-and-Unfold strategy, facilitating flexible per-mode encoding of high-order tensors via recursive unfold-encode-fold operations, effectively integrating tensor structural priors. Notably, MA-NTAE exhibits linear growth in computational complexity with tensor order and proportional growth with mode dimensions. Extensive experiments demonstrate MA-NTAE's performance advantages over standard AE and current tensor networks in compression and clustering tasks, which become increasingly pronounced for higher-order, higher-dimensional tensors.
Authors: Rui Liu, Haolin Zuo, Zheng Lian, Hongyu Yuan, Qi Fan
Abstract: Missing modalities have recently emerged as a critical research direction in multimodal emotion recognition (MER). Conventional approaches typically address this issue through missing modality reconstruction. However, these methods fail to account for variations in reconstruction difficulty across different samples, consequently limiting the model's ability to handle hard samples effectively. To overcome this limitation, we propose a novel Hardness-Aware Dynamic Curriculum Learning framework, termed HARDY-MER. Our framework operates in two key stages: first, it estimates the hardness level of each sample, and second, it strategically emphasizes hard samples during training to enhance model performance on these challenging instances. Specifically, we first introduce a Multi-view Hardness Evaluation mechanism that quantifies reconstruction difficulty by considering both Direct Hardness (modality reconstruction errors) and Indirect Hardness (cross-modal mutual information). Meanwhile, we introduce a Retrieval-based Dynamic Curriculum Learning strategy that dynamically adjusts the training curriculum by retrieving samples with similar semantic information and balancing the learning focus between easy and hard instances. Extensive experiments on benchmark datasets demonstrate that HARDY-MER consistently outperforms existing methods in missing-modality scenarios. Our code will be made publicly available at https://github.com/HARDY-MER/HARDY-MER.
Authors: Xiao Huang, Xu Liu, Enze Zhang, Tong Yu, Shuai Li
Abstract: Offline-to-online Reinforcement Learning (O2O RL) aims to perform online fine-tuning on an offline pre-trained policy to minimize costly online interactions. Existing work used offline datasets to generate data that conform to the online data distribution for data augmentation. However, generated data still exhibits a gap with the online data, limiting overall performance. To address this, we propose a new data augmentation approach, Classifier-Free Diffusion Generation (CFDG). Without introducing additional classifier training overhead, CFDG leverages classifier-free guidance diffusion to significantly enhance the generation quality of offline and online data with different distributions. Additionally, it employs a reweighting method to enable more generated data to align with the online data, enhancing performance while maintaining the agent's stability. Experimental results show that CFDG outperforms replaying the two data types or using a standard diffusion model to generate new data. Our method is versatile and can be integrated with existing offline-to-online RL algorithms. By implementing CFDG to popular methods IQL, PEX and APL, we achieve a notable 15% average improvement in empirical performance on the D4RL benchmark such as MuJoCo and AntMaze.
Authors: Brendan R. Hogan, Will Brown, Adel Boyarsky, Anderson Schneider, Yuriy Nevmyvaka
Abstract: Even though large language models are becoming increasingly capable, it is still unreasonable to expect them to excel at tasks that are under-represented on the Internet. Leveraging LLMs for specialized applications, particularly in niche programming languages and private domains, remains challenging and largely unsolved. In this work, we address this gap by presenting a comprehensive, open-source approach for adapting LLMs to the Q programming language, a popular tool in quantitative finance that is much less present on the Internet compared to Python, C, Java, and other ``mainstream" languages and is therefore not a strong suit of general-purpose AI models. We introduce a new Leetcode style evaluation dataset for Q, benchmark major frontier models on the dataset, then do pretraining, supervised fine tuning, and reinforcement learning to train a suite of reasoning and non-reasoning models based on the Qwen-2.5 series, spanning five parameter sizes (1.5B, 3B, 7B, 14B, 32B). Our best model achieves a pass@1 accuracy of 59 percent on our Q benchmark, surpassing the best-performing frontier model, Claude Opus-4 by 29.5 percent. Additionally, all models, even our 1.5B model, outperform GPT-4.1 on this task. In addition to releasing models, code, and data, we provide a detailed blueprint for dataset construction, model pretraining, supervised fine-tuning, and reinforcement learning. Our methodology is broadly applicable, and we discuss how these techniques can be extended to other tasks, including those where evaluation may rely on soft or subjective signals.
Authors: Ishwar Balappanawar, Venkata Hasith Vattikuti, Greta Kintzley, Ronan Azimi-Mancel, Satvik Golechha
Abstract: Detecting hidden behaviors in neural networks poses a significant challenge due to minimal prior knowledge and potential adversarial obfuscation. We explore this problem by framing detection as an adversarial game between two teams: the red team trains two similar models, one trained solely on benign data and the other trained on data containing hidden harmful behavior, with the performance of both being nearly indistinguishable on the benign dataset. The blue team, with limited to no information about the harmful behaviour, tries to identify the compromised model. We experiment using CNNs and try various blue team strategies, including Gaussian noise analysis, model diffing, integrated gradients, and adversarial attacks under different levels of hints provided by the red team. Results show high accuracy for adversarial-attack-based methods (100\% correct prediction, using hints), which is very promising, whilst the other techniques yield more varied performance. During our LLM-focused rounds, we find that there are not many parallel methods that we could apply from our study with CNNs. Instead, we find that effective LLM auditing methods require some hints about the undesired distribution, which can then used in standard black-box and open-weight methods to probe the models further and reveal their misalignment. We open-source our auditing games (with the model and data) and hope that our findings contribute to designing better audits.
Authors: Aleksandar Todorov, Juan Cardenas-Cartagena, Rafael F. Cunha, Marco Zullich, Matthia Sabatelli
Abstract: Plasticity loss, a diminishing capacity to adapt as training progresses, is a critical challenge in deep reinforcement learning. We examine this issue in multi-task reinforcement learning (MTRL), where higher representational flexibility is crucial for managing diverse and potentially conflicting task demands. We systematically explore how sparsification methods, particularly Gradual Magnitude Pruning (GMP) and Sparse Evolutionary Training (SET), enhance plasticity and consequently improve performance in MTRL agents. We evaluate these approaches across distinct MTRL architectures (shared backbone, Mixture of Experts, Mixture of Orthogonal Experts) on standardized MTRL benchmarks, comparing against dense baselines, and a comprehensive range of alternative plasticity-inducing or regularization methods. Our results demonstrate that both GMP and SET effectively mitigate key indicators of plasticity degradation, such as neuron dormancy and representational collapse. These plasticity improvements often correlate with enhanced multi-task performance, with sparse agents frequently outperforming dense counterparts and achieving competitive results against explicit plasticity interventions. Our findings offer insights into the interplay between plasticity, network sparsity, and MTRL designs, highlighting dynamic sparsification as a robust but context-sensitive tool for developing more adaptable MTRL systems.
Authors: Anthony Bellotti, Xindi Zhao
Abstract: Conformal predictors are machine learning algorithms developed in the 1990's by Gammerman, Vovk, and their research team, to provide set predictions with guaranteed confidence level. Over recent years, they have grown in popularity and have become a mainstream methodology for uncertainty quantification in the machine learning community. From its beginning, there was an understanding that they enable reliable machine learning with well-calibrated uncertainty quantification. This makes them extremely beneficial for developing trustworthy AI, a topic that has also risen in interest over the past few years, in both the AI community and society more widely. In this article, we review the potential for conformal prediction to contribute to trustworthy AI beyond its marginal validity property, addressing problems such as generalization risk and AI governance. Experiments and examples are also provided to demonstrate its use as a well-calibrated predictor and for bias identification and mitigation.
Authors: Shichao Ma, Zhengyang Zhou, Qihe Huang, Binwu Wang, Kuo Yang, Huan Li, Yang Wang
Abstract: Time series forecasting has become increasingly important to empower diverse applications with streaming data. Zero-shot time-series forecasting (ZSF), particularly valuable in data-scarce scenarios, such as domain transfer or forecasting under extreme conditions, is difficult for traditional models to deal with. While time series pre-trained models (TSPMs) have demonstrated strong performance in ZSF, they often lack mechanisms to dynamically incorporate external knowledge. Fortunately, emerging retrieval-augmented generation (RAG) offers a promising path for injecting such knowledge on demand, yet they are rarely integrated with TSPMs. To leverage the strengths of both worlds, we introduce RAG into TSPMs to enhance zero-shot time series forecasting. In this paper, we propose QuiZSF (Quick Zero-Shot Time Series Forecaster), a lightweight and modular framework that couples efficient retrieval with representation learning and model adaptation for ZSF. Specifically, we construct a hierarchical tree-structured ChronoRAG Base (CRB) for scalable time-series storage and domain-aware retrieval, introduce a Multi-grained Series Interaction Learner (MSIL) to extract fine- and coarse-grained relational features, and develop a dual-branch Model Cooperation Coherer (MCC) that aligns retrieved knowledge with two kinds of TSPMs: Non-LLM based and LLM based. Compared with contemporary baselines, QuiZSF, with Non-LLM based and LLM based TSPMs as base model, respectively, ranks Top1 in 75% and 87.5% of prediction settings, while maintaining high efficiency in memory and inference time.
Authors: Lishi Zuo, Man-Wai Mak, Lu Yi, Youzhi Tu
Abstract: Medical diagnosis might fail due to bias. In this work, we identified class-feature bias, which refers to models' potential reliance on features that are strongly correlated with only a subset of classes, leading to biased performance and poor generalization on other classes. We aim to train a class-unbiased model (Cls-unbias) that mitigates both class imbalance and class-feature bias simultaneously. Specifically, we propose a class-wise inequality loss which promotes equal contributions of classification loss from positive-class and negative-class samples. We propose to optimize a class-wise group distributionally robust optimization objective-a class-weighted training objective that upweights underperforming classes-to enhance the effectiveness of the inequality loss under class imbalance. Through synthetic and real-world datasets, we empirically demonstrate that class-feature bias can negatively impact model performance. Our proposed method effectively mitigates both class-feature bias and class imbalance, thereby improving the model's generalization ability.
Authors: Lixuan He, Jie Feng, Yong Li
Abstract: Large Language Models (LLMs) are typically fine-tuned for reasoning tasks through a two-stage pipeline of Supervised Fine-Tuning (SFT) followed by Reinforcement Learning (RL), a process fraught with catastrophic forgetting and suboptimal trade-offs between imitation and exploration. Recent single-stage methods attempt to unify SFT and RL using heuristics, but lack a principled mechanism for dynamically balancing the two paradigms. In this paper, we reframe this challenge through the theoretical lens of \textbf{implicit rewards}, viewing SFT and RL not as distinct methods but as complementary reward signals. We introduce \textbf{Adaptive Meta Fine-Tuning (AMFT)}, a novel single-stage algorithm that learns the optimal balance between SFT's implicit, path-level reward and RL's explicit, outcome-based reward. The core of AMFT is a \textbf{meta-gradient adaptive weight controller} that treats the SFT-RL balance as a learnable parameter, dynamically optimizing it to maximize long-term task performance. This forward-looking approach, regularized by policy entropy for stability, autonomously discovers an effective training curriculum. We conduct a comprehensive evaluation on challenging benchmarks spanning mathematical reasoning, abstract visual reasoning (General Points), and vision-language navigation (V-IRL). AMFT consistently establishes a new state-of-the-art and demonstrats superior generalization on out-of-distribution (OOD) tasks. Ablation studies and training dynamic analysis confirm that the meta-learning controller is crucial for AMFT's stability, sample efficiency, and performance, offering a more principled and effective paradigm for LLM alignment. Our codes are open-sourced via https://github.com/hlxtsyj/AMFT.
Authors: Shiwei Li, Xiandi Luo, Haozhao Wang, Xing Tang, Ziqiang Cui, Dugang Liu, Yuhua Li, Xiuqiang He, Ruixuan Li
Abstract: Low-rank adaptation (LoRA) is a parameter-efficient fine-tuning (PEFT) method widely used in large language models (LLMs). It approximates the update of a pretrained weight matrix $W\in\mathbb{R}^{m\times n}$ by the product of two low-rank matrices, $BA$, where $A \in\mathbb{R}^{r\times n}$ and $B\in\mathbb{R}^{m\times r} (r\ll\min\{m,n\})$. Increasing the dimension $r$ can raise the rank of LoRA weights (i.e., $BA$), which typically improves fine-tuning performance but also significantly increases the number of trainable parameters. In this paper, we propose Block Diversified Low-Rank Adaptation (BoRA), which improves the rank of LoRA weights with a small number of additional parameters. Specifically, BoRA treats the product $BA$ as a block matrix multiplication, where $A$ and $B$ are partitioned into $b$ blocks along the columns and rows, respectively (i.e., $A=[A_1,\dots,A_b]$ and $B=[B_1,\dots,B_b]^\top$). Consequently, the product $BA$ becomes the concatenation of the block products $B_iA_j$ for $i,j\in[b]$. To enhance the diversity of different block products, BoRA introduces a unique diagonal matrix $\Sigma_{i,j} \in \mathbb{R}^{r\times r}$ for each block multiplication, resulting in $B_i \Sigma_{i,j} A_j$. By leveraging these block-wise diagonal matrices, BoRA increases the rank of LoRA weights by a factor of $b$ while only requiring $b^2r$ additional parameters. Extensive experiments across multiple datasets and models demonstrate the superiority of BoRA, and ablation studies further validate its scalability.
Authors: Hiba Najjar, Bushra Alshbib, Andreas Dengel
Abstract: Remote sensing provides satellite data in diverse types and formats. The usage of multimodal learning networks exploits this diversity to improve model performance, except that the complexity of such networks comes at the expense of their interpretability. In this study, we explore how modalities can be leveraged through multitask learning to intrinsically explain model behavior. In particular, instead of additional inputs, we use certain modalities as additional targets to be predicted along with the main task. The success of this approach relies on the rich information content of satellite data, which remains as input modalities. We show how this modeling context provides numerous benefits: (1) in case of data scarcity, the additional modalities do not need to be collected for model inference at deployment, (2) the model performance remains comparable to the multimodal baseline performance, and in some cases achieves better scores, (3) prediction errors in the main task can be explained via the model behavior in the auxiliary task(s). We demonstrate the efficiency of our approach on three datasets, including segmentation, classification, and regression tasks. Code available at git.opendfki.de/hiba.najjar/mtl_explainability/.
Authors: Brooks Kinch, Benjamin Shaffer, Elizabeth Armstrong, Michael Meehan, John Hewson, Nathaniel Trask
Abstract: We present a framework for constructing real-time digital twins based on structure-preserving reduced finite element models conditioned on a latent variable Z. The approach uses conditional attention mechanisms to learn both a reduced finite element basis and a nonlinear conservation law within the framework of finite element exterior calculus (FEEC). This guarantees numerical well-posedness and exact preservation of conserved quantities, regardless of data sparsity or optimization error. The conditioning mechanism supports real-time calibration to parametric variables, allowing the construction of digital twins which support closed loop inference and calibration to sensor data. The framework interfaces with conventional finite element machinery in a non-invasive manner, allowing treatment of complex geometries and integration of learned models with conventional finite element techniques. Benchmarks include advection diffusion, shock hydrodynamics, electrostatics, and a complex battery thermal runaway problem. The method achieves accurate predictions on complex geometries with sparse data (25 LES simulations), including capturing the transition to turbulence and achieving real-time inference ~0.1s with a speedup of 3.1x10^8 relative to LES. An open-source implementation is available on GitHub.
Authors: Jiawei Zhang, Yifei Zhang, Baozhao Yi, Yao Ren, Qi Jiao, Hanyu Bai, Weiran Jiang, Ziyou Song
Abstract: Fast and reliable validation of novel designs in complex physical systems such as batteries is critical to accelerating technological innovation. However, battery research and development remain bottlenecked by the prohibitively high time and energy costs required to evaluate numerous new design candidates, particularly in battery prototyping and life testing. Despite recent progress in data-driven battery lifetime prediction, existing methods require labeled data of target designs to improve accuracy and cannot make reliable predictions until after prototyping, thus falling far short of the efficiency needed to enable rapid feedback for battery design. Here, we introduce Discovery Learning (DL), a scientific machine-learning paradigm that integrates active learning, physics-guided learning, and zero-shot learning into a human-like reasoning loop, drawing inspiration from learning theories in educational psychology. DL can learn from historical battery designs and actively reduce the need for prototyping, thus enabling rapid lifetime evaluation for unobserved material-design combinations without requiring additional data labeling. To test DL, we present 123 industrial-grade large-format lithium-ion pouch cells, spanning eight material-design combinations and diverse cycling protocols. Trained solely on public datasets of small-capacity cylindrical cells, DL achieves 7.2% test error in predicting the average cycle life under unknown device variability. This results in savings of 98% in time and 95% in energy compared to industrial practices. This work highlights the potential of uncovering insights from historical designs to inform and accelerate the development of next-generation battery technologies. DL represents a key advance toward efficient data-driven modeling and helps realize the promise of machine learning for accelerating scientific discovery and engineering innovation.
Authors: Chonghua Han, Yuan Yuan, Yukun Liu, Jingtao Ding, Jie Feng, Yong Li
Abstract: Human mobility prediction is vital for urban planning, transportation optimization, and personalized services. However, the inherent randomness, non-uniform time intervals, and complex patterns of human mobility, compounded by the heterogeneity introduced by varying city structures, infrastructure, and population densities, present significant challenges in modeling. Existing solutions often require training separate models for each city due to distinct spatial representations and geographic coverage. In this paper, we propose UniMove, a unified model for multi-city human mobility prediction, addressing two challenges: (1) constructing universal spatial representations for effective token sharing across cities, and (2) modeling heterogeneous mobility patterns from varying city characteristics. We propose a trajectory-location dual-tower architecture, with a location tower for universal spatial encoding and a trajectory tower for sequential mobility modeling. We also design MoE Transformer blocks to adaptively select experts to handle diverse movement patterns. Extensive experiments across multiple datasets from diverse cities demonstrate that UniMove truly embodies the essence of a unified model. By enabling joint training on multi-city data with mutual data enhancement, it significantly improves mobility prediction accuracy by over 10.2\%. UniMove represents a key advancement toward realizing a true foundational model with a unified architecture for human mobility. We release the implementation at https://github.com/tsinghua-fib-lab/UniMove/.
Authors: Vojtech Halenka, Ole-Christoffer Granmo, Lei Jiao, Per-Arne Andersen
Abstract: Feature Selection (FS) is crucial for improving model interpretability, reducing complexity, and sometimes for enhancing accuracy. The recently introduced Tsetlin machine (TM) offers interpretable clause-based learning, but lacks established tools for estimating feature importance. In this paper, we adapt and evaluate a range of FS techniques for TMs, including classical filter and embedded methods as well as post-hoc explanation methods originally developed for neural networks (e.g., SHAP and LIME) and a novel family of embedded scorers derived from TM clause weights and Tsetlin automaton (TA) states. We benchmark all methods across 12 datasets, using evaluation protocols, like Remove and Retrain (ROAR) strategy and Remove and Debias (ROAD), to assess causal impact. Our results show that TM-internal scorers not only perform competitively but also exploit the interpretability of clauses to reveal interacting feature patterns. Simpler TM-specific scorers achieve similar accuracy retention at a fraction of the computational cost. This study establishes the first comprehensive baseline for FS in TM and paves the way for developing specialized TM-specific interpretability techniques.
Authors: Helbert Paat, Guohao Shen
Abstract: Decision support systems are designed to assist human experts in classification tasks by providing conformal prediction sets derived from a pre-trained model. This human-AI collaboration has demonstrated enhanced classification performance compared to using either the model or the expert independently. In this study, we focus on the selection of instance-specific experts from a pool of multiple human experts, contrasting it with existing research that typically focuses on single-expert scenarios. We characterize the conditions under which multiple experts can benefit from the conformal sets. With the insight that only certain experts may be relevant for each instance, we explore the problem of subset selection and introduce a greedy algorithm that utilizes conformal sets to identify the subset of expert predictions that will be used in classifying an instance. This approach is shown to yield better performance compared to naive methods for human subset selection. Based on real expert predictions from the CIFAR-10H and ImageNet-16H datasets, our simulation study indicates that our proposed greedy algorithm achieves near-optimal subsets, resulting in improved classification performance among multiple experts.
Authors: Jianfei Wu, Wenmian Yang, Bingning Liu, Weijia Jia
Abstract: Time series forecasting is critical across various domains, such as weather, finance and real estate forecasting, as accurate forecasts support informed decision-making and risk mitigation. While recent deep learning models have improved predictive capabilities, they often overlook time-lagged cross-correlations between related sequences, which are crucial for capturing complex temporal relationships. To address this, we propose the Time-Lagged Cross-Correlations-based Sequence Prediction framework (TLCCSP), which enhances forecasting accuracy by effectively integrating time-lagged cross-correlated sequences. TLCCSP employs the Sequence Shifted Dynamic Time Warping (SSDTW) algorithm to capture lagged correlations and a contrastive learning-based encoder to efficiently approximate SSDTW distances. Experimental results on weather, finance and real estate time series datasets demonstrate the effectiveness of our framework. On the weather dataset, SSDTW reduces mean squared error (MSE) by 16.01% compared with single-sequence methods, while the contrastive learning encoder (CLE) further decreases MSE by 17.88%. On the stock dataset, SSDTW achieves a 9.95% MSE reduction, and CLE reduces it by 6.13%. For the real estate dataset, SSDTW and CLE reduce MSE by 21.29% and 8.62%, respectively. Additionally, the contrastive learning approach decreases SSDTW computational time by approximately 99%, ensuring scalability and real-time applicability across multiple time series forecasting tasks.
Authors: Antonio Guillen-Perez
Abstract: Learning robust driving policies from large-scale, real-world datasets is a central challenge in autonomous driving, as online data collection is often unsafe and impractical. While Behavioral Cloning (BC) offers a straightforward approach to imitation learning, policies trained with BC are notoriously brittle and suffer from compounding errors in closed-loop execution. This work presents a comprehensive pipeline and a comparative study to address this limitation. We first develop a series of increasingly sophisticated BC baselines, culminating in a Transformer-based model that operates on a structured, entity-centric state representation. While this model achieves low imitation loss, we show that it still fails in long-horizon simulations. We then demonstrate that by applying a state-of-the-art Offline Reinforcement Learning algorithm, Conservative Q-Learning (CQL), to the same data and architecture, we can learn a significantly more robust policy. Using a carefully engineered reward function, the CQL agent learns a conservative value function that enables it to recover from minor errors and avoid out-of-distribution states. In a large-scale evaluation on 1,000 unseen scenarios from the Waymo Open Motion Dataset, our final CQL agent achieves a 3.2x higher success rate and a 7.4x lower collision rate than the strongest BC baseline, proving that an offline RL approach is critical for learning robust, long-horizon driving policies from static expert data.
Authors: Tiantian He, Keyue Jiang, An Zhao, Anna Schroder, Elinor Thompson, Sonja Soskic, Frederik Barkhof, Daniel C. Alexander
Abstract: The long-term progression of neurodegenerative diseases is commonly conceptualized as a spatiotemporal diffusion process that consists of a graph diffusion process across the structural brain connectome and a localized reaction process within brain regions. However, modeling this progression remains challenging due to 1) the scarcity of longitudinal data obtained through irregular and infrequent subject visits and 2) the complex interplay of pathological mechanisms across brain regions and disease stages, where traditional models assume fixed mechanisms throughout disease progression. To address these limitations, we propose a novel stage-aware Mixture of Experts (MoE) framework that explicitly models how different contributing mechanisms dominate at different disease stages through time-dependent expert weighting.Data-wise, we utilize an iterative dual optimization method to properly estimate the temporal position of individual observations, constructing a co hort-level progression trajectory from irregular snapshots. Model-wise, we enhance the spatial component with an inhomogeneous graph neural diffusion model (IGND) that allows diffusivity to vary based on node states and time, providing more flexible representations of brain networks. We also introduce a localized neural reaction module to capture complex dynamics beyond standard processes.The resulting IGND-MoE model dynamically integrates these components across temporal states, offering a principled way to understand how stage-specific pathological mechanisms contribute to progression. The stage-wise weights yield novel clinical insights that align with literature, suggesting that graph-related processes are more influential at early stages, while other unknown physical processes become dominant later on.
Authors: Yangguang He, Wenhao Li, Minzhe Li, Juan Zhang, Xiangfeng Wang, Bo Jin
Abstract: Learning-based filtering has demonstrated strong performance in non-linear dynamical systems, particularly when the statistics of noise are unknown. However, in real-world deployments, environmental factors, such as changing wind conditions or electromagnetic interference, can induce unobserved noise-statistics drift, leading to substantial degradation of learning-based methods. To address this challenge, we propose OTAKNet, the first online solution to noise-statistics drift within learning-based adaptive Kalman filtering. Unlike existing learning-based methods that perform offline fine-tuning using batch pointwise matching over entire trajectories, OTAKNet establishes a connection between the state estimate and the drift via one-step predictive measurement likelihood, and addresses it using optimal transport. This leverages OT's geometry - aware cost and stable gradients to enable fully online adaptation without ground truth labels or retraining. We compare OTAKNet against classical model-based adaptive Kalman filtering and offline learning-based filtering. The performance is demonstrated on both synthetic and real-world NCLT datasets, particularly under limited training data.
Authors: Ziqi Zhang, Ali Shahin Shamsabadi, Hanxiao Lu, Yifeng Cai, Hamed Haddadi
Abstract: Recent advances in Knowledge Distillation (KD) aim to mitigate the high computational demands of Large Language Models (LLMs) by transferring knowledge from a large ''teacher'' to a smaller ''student'' model. However, students may inherit the teacher's privacy when the teacher is trained on private data. In this work, we systematically characterize and investigate membership and memorization privacy risks inherent in six LLM KD techniques. Using instruction-tuning settings that span seven NLP tasks, together with three teacher model families (GPT-2, LLAMA-2, and OPT), and various size student models, we demonstrate that all existing LLM KD approaches carry membership and memorization privacy risks from the teacher to its students. However, the extent of privacy risks varies across different KD techniques. We systematically analyse how key LLM KD components (KD objective functions, student training data and NLP tasks) impact such privacy risks. We also demonstrate a significant disagreement between memorization and membership privacy risks of LLM KD techniques. Finally, we characterize per-block privacy risk and demonstrate that the privacy risk varies across different blocks by a large margin.
Authors: Stanley Ngugi
Abstract: Large Language Models (LLMs) struggle with dynamic knowledge updates, especially when new information conflicts with deeply embedded facts. Such conflicting factual edits often lead to two critical issues: resistance to adopting the new fact and severe catastrophic forgetting of unrelated knowledge. This paper introduces and evaluates a novel "unlearn-then-learn" strategy for precise knowledge editing in LLMs, leveraging the parameter-efficient fine-tuning (PEFT) technique, Infused Adapter by Inhibiting and Amplifying Inner Activations ($IA^3$). Crucially, this two-stage approach is powered by an initial circuit localization phase that identifies and targets the specific internal components responsible for encoding the conflicting fact. Through a rigorous experimental methodology on microsoft/Phi-3-mini-4k-instruct, we demonstrate that this mechanistically informed two-stage approach achieves near-perfect accuracy (98.50%) for the new, modulated fact while simultaneously effectively suppressing the original conflicting fact (96.00% forget rate). Critically, our strategy exhibits unprecedented localization (72.00% F_control accuracy), dramatically mitigating catastrophic forgetting observed in direct fine-tuning approaches (which showed as low as ~20% F_control accuracy), a direct benefit of our targeted interpretability-guided intervention. Furthermore, qualitative analysis reveals a nuanced mechanism of "soft forgetting," where original knowledge is suppressed from default retrieval but remains latent and conditionally accessible, enhancing model safety and control. These findings represent a significant advancement towards precise, localized, and safe knowledge management in compact LLMs.
Authors: N Harshit, K Mounvik
Abstract: In applied machine learning, concept drift, which is either gradual or abrupt changes in data distribution, can significantly reduce model performance. Typical detection methods,such as statistical tests or reconstruction-based models,are generally reactive and not very sensitive to early detection. Our study proposes a hybrid framework consisting of Transformers and Autoencoders to model complex temporal dynamics and provide online drift detection. We create a distinct Trust Score methodology, which includes signals on (1) statistical and reconstruction-based drift metrics, more specifically, PSI, JSD, Transformer-AE error, (2) prediction uncertainty, (3) rules violations, and (4) trend of classifier error aligned with the combined metrics defined by the Trust Score. Using a time sequenced airline passenger data set with synthetic drift, our proposed model allows for a better detection of drift using as a whole and at different detection thresholds for both sensitivity and interpretability compared to baseline methods and provides a strong pipeline for drift detection in real time for applied machine learning. We evaluated performance using a time-sequenced airline passenger dataset having the gradually injected stimulus of drift in expectations,e.g. permuted ticket prices in later batches, broken into 10 time segments [1].In the data, our results support that the Transformation-Autoencoder detected drift earlier and with more sensitivity than the autoencoders commonly used in the literature, and provided improved modeling over more error rates and logical violations. Therefore, a robust framework was developed to reliably monitor concept drift.
Authors: Yang Cao, Yubin Chen, Zhao Song, Jiahao Zhang
Abstract: Generative modelling has seen significant advances through simulation-free paradigms such as Flow Matching, and in particular, the MeanFlow framework, which replaces instantaneous velocity fields with average velocities to enable efficient single-step sampling. In this work, we introduce a theoretical study on Second-Order MeanFlow, a novel extension that incorporates average acceleration fields into the MeanFlow objective. We first establish the feasibility of our approach by proving that the average acceleration satisfies a generalized consistency condition analogous to first-order MeanFlow, thereby supporting stable, one-step sampling and tractable loss functions. We then characterize its expressivity via circuit complexity analysis, showing that under mild assumptions, the Second-Order MeanFlow sampling process can be implemented by uniform threshold circuits within the $\mathsf{TC}^0$ class. Finally, we derive provably efficient criteria for scalable implementation by leveraging fast approximate attention computations: we prove that attention operations within the Second-Order MeanFlow architecture can be approximated to within $1/\mathrm{poly}(n)$ error in time $n^{2+o(1)}$. Together, these results lay the theoretical foundation for high-order flow matching models that combine rich dynamics with practical sampling efficiency.
Authors: Yiran Huang, Amirhossein Nouranizadeh, Christine Ahrends, Mengjia Xu
Abstract: Functional Magnetic Resonance Imaging (fMRI) is an imaging technique widely used to study human brain activity. fMRI signals in areas across the brain transiently synchronise and desynchronise their activity in a highly structured manner, even when an individual is at rest. These functional connectivity dynamics may be related to behaviour and neuropsychiatric disease. To model these dynamics, temporal brain connectivity representations are essential, as they reflect evolving interactions between brain regions and provide insight into transient neural states and network reconfigurations. However, conventional graph neural networks (GNNs) often struggle to capture long-range temporal dependencies in dynamic fMRI data. To address this challenge, we propose BrainATCL, an unsupervised, nonparametric framework for adaptive temporal brain connectivity learning, enabling functional link prediction and age estimation. Our method dynamically adjusts the lookback window for each snapshot based on the rate of newly added edges. Graph sequences are subsequently encoded using a GINE-Mamba2 backbone to learn spatial-temporal representations of dynamic functional connectivity in resting-state fMRI data of 1,000 participants from the Human Connectome Project. To further improve spatial modeling, we incorporate brain structure and function-informed edge attributes, i.e., the left/right hemispheric identity and subnetwork membership of brain regions, enabling the model to capture biologically meaningful topological patterns. We evaluate our BrainATCL on two tasks: functional link prediction and age estimation. The experimental results demonstrate superior performance and strong generalization, including in cross-session prediction scenarios.
Authors: Atakan Azakli, Bernd Stelzer
Abstract: In this work, we propose a new machine learning (ML) methodology to obtain more precise predictions for some parameters of interest in a given hypotheses testing problem. Our proposed method also allows ML models to have more discriminative power in cases where it is extremely challenging for state-of-the-art classifiers to have any level of accurate predictions. This method can also allow us to systematically decrease the error from ML models in their predictions. In this paper, we provide a mathematical motivation why Multiple Instance Learning (MIL) would have more predictive power over their single-instance counterparts. We support our theoretical claims by analyzing the behavior of the MIL models through their scaling behaviors with respect to the number of instances on which the model makes predictions. As a concrete application, we constrain Wilson coefficients of the Standard Model Effective Field Theory (SMEFT) using kinematic information from subatomic particle collision events at the Large Hadron Collider (LHC). We show that under certain circumstances, it might be possible to extract the theoretical maximum Fisher Information latent in a dataset.
Authors: Peyman Baghershahi, Gregoire Fournier, Pranav Nyati, Sourav Medya
Abstract: Graph Neural Networks (GNNs) have emerged as powerful tools for learning over structured data, including text-attributed graphs, which are common in domains such as citation networks, social platforms, and knowledge graphs. GNNs are not inherently interpretable and thus, many explanation methods have been proposed. However, existing explanation methods often struggle to generate interpretable, fine-grained rationales, especially when node attributes include rich natural language. In this work, we introduce LOGIC, a lightweight, post-hoc framework that uses large language models (LLMs) to generate faithful and interpretable explanations for GNN predictions. LOGIC projects GNN node embeddings into the LLM embedding space and constructs hybrid prompts that interleave soft prompts with textual inputs from the graph structure. This enables the LLM to reason about GNN internal representations and produce natural language explanations along with concise explanation subgraphs. Our experiments across four real-world TAG datasets demonstrate that LOGIC achieves a favorable trade-off between fidelity and sparsity, while significantly improving human-centric metrics such as insightfulness. LOGIC sets a new direction for LLM-based explainability in graph learning by aligning GNN internals with human reasoning.
Authors: Zhihao Xue, Yun Zi, Nia Qi, Ming Gong, Yujun Zou
Abstract: This paper proposes a spatiotemporal graph neural network-based performance prediction algorithm to address the challenge of forecasting performance fluctuations in distributed backend systems with multi-level service call structures. The method abstracts system states at different time slices into a sequence of graph structures. It integrates the runtime features of service nodes with the invocation relationships among services to construct a unified spatiotemporal modeling framework. The model first applies a graph convolutional network to extract high-order dependency information from the service topology. Then it uses a gated recurrent network to capture the dynamic evolution of performance metrics over time. A time encoding mechanism is also introduced to enhance the model's ability to represent non-stationary temporal sequences. The architecture is trained in an end-to-end manner, optimizing the multi-layer nested structure to achieve high-precision regression of future service performance metrics. To validate the effectiveness of the proposed method, a large-scale public cluster dataset is used. A series of multi-dimensional experiments are designed, including variations in time windows and concurrent load levels. These experiments comprehensively evaluate the model's predictive performance and stability. The experimental results show that the proposed model outperforms existing representative methods across key metrics such as MAE, RMSE, and R2. It maintains strong robustness under varying load intensities and structural complexities. These results demonstrate the model's practical potential for backend service performance management tasks.
Authors: Zhengran Ji, Boyuan Chen
Abstract: Training reinforcement learning agents with human feedback is crucial when task objectives are difficult to specify through dense reward functions. While prior methods rely on offline trajectory comparisons to elicit human preferences, such data is unavailable in online learning scenarios where agents must adapt on the fly. Recent approaches address this by collecting real-time scalar feedback to guide agent behavior and train reward models for continued learning after human feedback becomes unavailable. However, scalar feedback is often noisy and inconsistent, limiting the accuracy and generalization of learned rewards. We propose Pref-GUIDE, a framework that transforms real-time scalar feedback into preference-based data to improve reward model learning for continual policy training. Pref-GUIDE Individual mitigates temporal inconsistency by comparing agent behaviors within short windows and filtering ambiguous feedback. Pref-GUIDE Voting further enhances robustness by aggregating reward models across a population of users to form consensus preferences. Across three challenging environments, Pref-GUIDE significantly outperforms scalar-feedback baselines, with the voting variant exceeding even expert-designed dense rewards. By reframing scalar feedback as structured preferences with population feedback, Pref-GUIDE offers a scalable and principled approach for harnessing human input in online reinforcement learning.
Authors: Niranjana Arun Menon, Iqra Farooq, Yulong Li, Sara Ahmed, Yutong Xie, Muhammad Awais, Imran Razzak
Abstract: Cardiovascular disease (CVD) prediction remains a tremendous challenge due to its multifactorial etiology and global burden of morbidity and mortality. Despite the growing availability of genomic and electrophysiological data, extracting biologically meaningful insights from such high-dimensional, noisy, and sparsely annotated datasets remains a non-trivial task. Recently, LLMs has been applied effectively to predict structural variations in biological sequences. In this work, we explore the potential of fine-tuned LLMs to predict cardiac diseases and SNPs potentially leading to CVD risk using genetic markers derived from high-throughput genomic profiling. We investigate the effect of genetic patterns associated with cardiac conditions and evaluate how LLMs can learn latent biological relationships from structured and semi-structured genomic data obtained by mapping genetic aspects that are inherited from the family tree. By framing the problem as a Chain of Thought (CoT) reasoning task, the models are prompted to generate disease labels and articulate informed clinical deductions across diverse patient profiles and phenotypes. The findings highlight the promise of LLMs in contributing to early detection, risk assessment, and ultimately, the advancement of personalized medicine in cardiac care.
Authors: Lu Chenggang
Abstract: Semi-Nonnegative Matrix Factorization (semi-NMF) extends classical Nonnegative Matrix Factorization (NMF) by allowing the basis matrix to contain both positive and negative entries, making it suitable for decomposing data with mixed signs. However, most existing semi-NMF algorithms are iterative, non-convex, and prone to local minima. In this paper, we propose a novel method that yields a globally optimal solution to the semi-NMF problem under the Frobenius norm, through an orthogonal decomposition derived from the scatter matrix of the input data. We rigorously prove that our solution attains the global minimum of the reconstruction error. Furthermore, we demonstrate that when the input matrix is nonnegative, our method often achieves lower reconstruction error than standard NMF algorithms, although unfortunately the basis matrix may not satisfy nonnegativity. In particular, in low-rank cases such as rank 1 or 2, our solution reduces exactly to a nonnegative factorization, recovering the NMF structure. We validate our approach through experiments on both synthetic data and the UCI Wine dataset, showing that our method consistently outperforms existing NMF and semi-NMF methods in terms of reconstruction accuracy. These results confirm that our globally optimal, non-iterative formulation offers both theoretical guarantees and empirical advantages, providing a new perspective on matrix factorization in optimization and data analysis.
Authors: Yuandong Tan
Abstract: The alignment of large language models (LLMs) with human preferences is commonly achieved through Reinforcement Learning from Human Feedback (RLHF). Direct Preference Optimization (DPO) simplified this paradigm by establishing a direct mapping between the optimal policy and a reward function, eliminating the need for an explicit reward model. However, we argue that the DPO loss function is theoretically misaligned with its own derivation, as it promotes the indefinite maximization of a logits difference, which can lead to training instability and reward hacking. In this paper, we propose a novel loss function derived directly from the RLHF optimality condition. Our proposed loss targets a specific, finite value for the logits difference, which is dictated by the underlying reward, rather than its maximization. We provide a theoretical analysis, including a gradient-based comparison, to demonstrate that our method avoids the large gradients that plague DPO when the probability of dispreferred responses approaches zero. This inherent stability prevents reward hacking and leads to more effective alignment. We validate our approach by fine-tuning a Qwen2.5-7B model, showing significant win-rate improvements over a standard DPO baseline and achieving competitive performance against larger models like Llama-3.1-8B.
Authors: Yashwant Krishna Pagoti, Arunesh Sinha, Shamik Sural
Abstract: In Federated Learning (FL), multiple clients jointly train a machine learning model by sharing gradient information, instead of raw data, with a server over multiple rounds. To address the possibility of information leakage in spite of sharing only the gradients, Local Differential Privacy (LDP) is often used. In LDP, clients add a selective amount of noise to the gradients before sending the same to the server. Although such noise addition protects the privacy of clients, it leads to a degradation in global model accuracy. In this paper, we model this privacy-accuracy trade-off as a game, where the sever incentivizes the clients to add a lower degree of noise for achieving higher accuracy, while the clients attempt to preserve their privacy at the cost of a potential loss in accuracy. A token based incentivization mechanism is introduced in which the quantum of tokens credited to a client in an FL round is a function of the degree of perturbation of its gradients. The client can later access a newly updated global model only after acquiring enough tokens, which are to be deducted from its balance. We identify the players, their actions and payoff, and perform a strategic analysis of the game. Extensive experiments were carried out to study the impact of different parameters.
Authors: Vincent-Daniel Yun
Abstract: Low-precision training has become essential for reducing the computational and memory costs of large-scale deep learning. However, quantization of gradients introduces both magnitude shrinkage and additive noise, which can alter the convergence behavior of stochastic gradient descent (SGD). In this work, we study the convergence of SGD under a gradient shrinkage model, where each stochastic gradient is scaled by a factor $q_k \in (0,1]$ and perturbed by zero-mean quantization noise. We show that this shrinkage is equivalent to replacing the nominal stepsize $\mu_k$ with an effective stepsize $\mu_k q_k$, which slows convergence when $q_{\min} < 1$. Under standard smoothness and bounded-variance assumptions, we prove that low-precision SGD still converges, but at a reduced rate determined by $q_{\min}$, and with an increased asymptotic error floor due to quantization noise. We theoretically analyze how reduced numerical precision slows down training by modeling it as gradient shrinkage in the standard SGD convergence framework.
Authors: Chanakya Ekbote, Marco Bondaschi, Nived Rajaraman, Jason D. Lee, Michael Gastpar, Ashok Vardhan Makkuva, Paul Pu Liang
Abstract: In-context learning (ICL) is a hallmark capability of transformers, through which trained models learn to adapt to new tasks by leveraging information from the input context. Prior work has shown that ICL emerges in transformers due to the presence of special circuits called induction heads. Given the equivalence between induction heads and conditional k-grams, a recent line of work modeling sequential inputs as Markov processes has revealed the fundamental impact of model depth on its ICL capabilities: while a two-layer transformer can efficiently represent a conditional 1-gram model, its single-layer counterpart cannot solve the task unless it is exponentially large. However, for higher order Markov sources, the best known constructions require at least three layers (each with a single attention head) - leaving open the question: can a two-layer single-head transformer represent any kth-order Markov process? In this paper, we precisely address this and theoretically show that a two-layer transformer with one head per layer can indeed represent any conditional k-gram. Thus, our result provides the tightest known characterization of the interplay between transformer depth and Markov order for ICL. Building on this, we further analyze the learning dynamics of our two-layer construction, focusing on a simplified variant for first-order Markov chains, illustrating how effective in-context representations emerge during training. Together, these results deepen our current understanding of transformer-based ICL and illustrate how even shallow architectures can surprisingly exhibit strong ICL capabilities on structured sequence modeling tasks.
Authors: Jian Xu, Yican Liu, Qibin Zhao, John Paisley, Delu Zeng
Abstract: Learning stochastic functions from partially observed context-target pairs is a fundamental problem in probabilistic modeling. Traditional models like Gaussian Processes (GPs) face scalability issues with large datasets and assume Gaussianity, limiting their applicability. While Neural Processes (NPs) offer more flexibility, they struggle with capturing complex, multi-modal target distributions. Neural Diffusion Processes (NDPs) enhance expressivity through a learned diffusion process but rely solely on conditional signals in the denoising network, resulting in weak input coupling from an unconditional forward process and semantic mismatch at the diffusion endpoint. In this work, we propose Neural Bridge Processes (NBPs), a novel method for modeling stochastic functions where inputs x act as dynamic anchors for the entire diffusion trajectory. By reformulating the forward kernel to explicitly depend on x, NBP enforces a constrained path that strictly terminates at the supervised target. This approach not only provides stronger gradient signals but also guarantees endpoint coherence. We validate NBPs on synthetic data, EEG signal regression and image regression tasks, achieving substantial improvements over baselines. These results underscore the effectiveness of DDPM-style bridge sampling in enhancing both performance and theoretical consistency for structured prediction tasks.
Authors: Po-Han Lee, Yu-Cheng Lin, Chan-Tung Ku, Chan Hsu, Pei-Cing Huang, Ping-Hsun Wu, Yihuang Kang
Abstract: Estimating individualized treatment effects from observational data presents a persistent challenge due to unmeasured confounding and structural bias. Causal Machine Learning (causal ML) methods, such as causal trees and doubly robust estimators, provide tools for estimating conditional average treatment effects. These methods have limited effectiveness in complex real-world environments due to the presence of latent confounders or those described in unstructured formats. Moreover, reliance on domain experts for confounder identification and rule interpretation introduces high annotation cost and scalability concerns. In this work, we proposed Large Language Model-based agents for automated confounder discovery and subgroup analysis that integrate agents into the causal ML pipeline to simulate domain expertise. Our framework systematically performs subgroup identification and confounding structure discovery by leveraging the reasoning capabilities of LLM-based agents, which reduces human dependency while preserving interpretability. Experiments on real-world medical datasets show that our proposed approach enhances treatment effect estimation robustness by narrowing confidence intervals and uncovering unrecognized confounding biases. Our findings suggest that LLM-based agents offer a promising path toward scalable, trustworthy, and semantically aware causal inference.
Authors: Ananda Prakash Verma
Abstract: We present EDGE, a general-purpose, misconception-aware adaptive learning framework composed of four stages: Evaluate (ability and state estimation), Diagnose (posterior infer-ence of misconceptions), Generate (counterfactual item synthesis), and Exercise (index-based retrieval scheduling). EDGE unifies psychometrics (IRT/Bayesian state space models), cog-nitive diagnostics (misconception discovery from distractor patterns and response latencies), contrastive item generation (minimal perturbations that invalidate learner shortcuts while pre-serving psychometric validity), and principled scheduling (a restless bandit approximation to spaced retrieval). We formalize a composite readiness metric, EdgeScore, prove its monotonicity and Lipschitz continuity, and derive an index policy that is near-optimal under mild assumptions on forgetting and learning gains. We further establish conditions under which counterfactual items provably reduce the posterior probability of a targeted misconception faster than standard practice. The paper focuses on theory and implementable pseudocode; empirical study is left to future work.
Authors: Chu Zhao, Eneng Yang, Yizhou Dang, Jianzhe Zhao, Guibing Guo, Xingwei Wang
Abstract: Heuristic negative sampling enhances recommendation performance by selecting negative samples of varying hardness levels from predefined candidate pools to guide the model toward learning more accurate decision boundaries. However, our empirical and theoretical analyses reveal that unobserved environmental confounders (e.g., exposure or popularity biases) in candidate pools may cause heuristic sampling methods to introduce false hard negatives (FHNS). These misleading samples can encourage the model to learn spurious correlations induced by such confounders, ultimately compromising its generalization ability under distribution shifts. To address this issue, we propose a novel method named Causal Negative Sampling via Diffusion (CNSDiff). By synthesizing negative samples in the latent space via a conditional diffusion process, CNSDiff avoids the bias introduced by predefined candidate pools and thus reduces the likelihood of generating FHNS. Moreover, it incorporates a causal regularization term to explicitly mitigate the influence of environmental confounders during the negative sampling process, leading to robust negatives that promote out-of-distribution (OOD) generalization. Comprehensive experiments under four representative distribution shift scenarios demonstrate that CNSDiff achieves an average improvement of 13.96% across all evaluation metrics compared to state-of-the-art baselines, verifying its effectiveness and robustness in OOD recommendation tasks.
Authors: Soumen Pachal, Mizhaan Prajit Maniyar, Prashanth L. A
Abstract: We consider the problem of risk-sensitive control in a reinforcement learning (RL) framework. In particular, we aim to find a risk-optimal policy by maximizing the distortion riskmetric (DRM) of the discounted reward in a finite horizon Markov decision process (MDP). DRMs are a rich class of risk measures that include several well-known risk measures as special cases. We derive a policy Hessian theorem for the DRM objective using the likelihood ratio method. Using this result, we propose a natural DRM Hessian estimator from sample trajectories of the underlying MDP. Next, we present a cubic-regularized policy Newton algorithm for solving this problem in an on-policy RL setting using estimates of the DRM gradient and Hessian. Our proposed algorithm is shown to converge to an $\epsilon$-second-order stationary point ($\epsilon$-SOSP) of the DRM objective, and this guarantee ensures the escaping of saddle points. The sample complexity of our algorithms to find an $ \epsilon$-SOSP is $\mathcal{O}(\epsilon^{-3.5})$. Our experiments validate the theoretical findings. To the best of our knowledge, our is the first work to present convergence to an $\epsilon$-SOSP of a risk-sensitive objective, while existing works in the literature have either shown convergence to a first-order stationary point of a risk-sensitive objective, or a SOSP of a risk-neutral one.
Authors: Bartlomiej Chybowski, Shima Abdullateef, Hollan Haule, Alfredo Gonzalez-Sulser, Javier Escudero
Abstract: Reliable seizure detection is critical for diagnosing and managing epilepsy, yet clinical workflows remain dependent on time-consuming manual EEG interpretation. While machine learning has shown promise, existing approaches often rely on dataset-specific optimisations, limiting their real-world applicability and reproducibility. Here, we introduce an innovative, open-source machine-learning framework that enables robust and generalisable seizure detection across varied clinical datasets. We evaluate our approach on two publicly available EEG datasets that differ in patient populations and electrode configurations. To enhance robustness, the framework incorporates an automated pre-processing pipeline to standardise data and a majority voting mechanism, in which multiple models independently assess each second of EEG before reaching a final decision. We train, tune, and evaluate models within each dataset, assessing their cross-dataset transferability. Our models achieve high within-dataset performance (AUC 0.904+/-0.059 for CHB-MIT and 0.864+/-0.060 for TUSZ) and demonstrate strong generalisation across datasets despite differences in EEG setups and populations (AUC 0.615+/-0.039 for models trained on CHB-MIT and tested on TUSZ and 0.762+/-0.175 in the reverse case) without any post-processing. Furthermore, a mild post-processing improved the within-dataset results to 0.913+/-0.064 and 0.867+/-0.058 and cross-dataset results to 0.619+/-0.036 and 0.768+/-0.172. These results underscore the potential of, and essential considerations for, deploying our framework in diverse clinical settings. By making our methodology fully reproducible, we provide a foundation for advancing clinically viable, dataset-agnostic seizure detection systems. This approach has the potential for widespread adoption, complementing rather than replacing expert interpretation, and accelerating clinical integration.
Authors: Hongbo Zhu, Angelo Cangelosi
Abstract: The goal of data attribution is to trace the model's predictions through the learning algorithm and back to its training data. thereby identifying the most influential training samples and understanding how the model's behavior leads to particular predictions. Understanding how individual training examples influence a model's predictions is fundamental for machine learning interpretability, data debugging, and model accountability. Influence functions, originating from robust statistics, offer an efficient, first-order approximation to estimate the impact of marginally upweighting or removing a data point on a model's learned parameters and its subsequent predictions, without the need for expensive retraining. This paper comprehensively reviews the data attribution capability of influence functions in deep learning. We discuss their theoretical foundations, recent algorithmic advances for efficient inverse-Hessian-vector product estimation, and evaluate their effectiveness for data attribution and mislabel detection. Finally, highlighting current challenges and promising directions for unleashing the huge potential of influence functions in large-scale, real-world deep learning scenarios.
Authors: Lin-Han Jia, Si-Yu Han, Wen-Chao Hu, Jie-Jing Shao, Wen-Da Wei, Zhi Zhou, Lan-Zhe Guo, Yu-Feng Li
Abstract: Neuro-symbolic (Nesy) learning improves the target task performance of models by enabling them to satisfy knowledge, while semi/self-supervised learning (SSL) improves the target task performance by designing unsupervised pretext tasks for unlabeled data to make models satisfy corresponding assumptions. We extend the Nesy theory based on reliable knowledge to the scenario of unreliable knowledge (i.e., assumptions), thereby unifying the theoretical frameworks of SSL and Nesy. Through rigorous theoretical analysis, we demonstrate that, in theory, the impact of pretext tasks on target performance hinges on three factors: knowledge learnability with respect to the model, knowledge reliability with respect to the data, and knowledge completeness with respect to the target. We further propose schemes to operationalize these theoretical metrics, and thereby develop a method that can predict the effectiveness of pretext tasks in advance. This will change the current status quo in practical applications, where the selections of unsupervised tasks are heuristic-based rather than theory-based, and it is difficult to evaluate the rationality of unsupervised pretext task selection before testing the model on the target task. In experiments, we verify a high correlation between the predicted performance-estimated using minimal data-and the actual performance achieved after large-scale semi-supervised or self-supervised learning, thus confirming the validity of the theory and the effectiveness of the evaluation method.
Authors: Tuo Zhang, Ning Li, Xin Yuan, Wenchao Xu, Quan Chen, Song Guo, Haijun Zhang
Abstract: With the breakthrough progress of large language models (LLMs) in natural language processing and multimodal tasks, efficiently deploying them on resource-constrained edge devices has become a critical challenge. The Mixture of Experts (MoE) architecture enhances model capacity through sparse activation, but faces two major difficulties in practical deployment: (1) The presence of numerous outliers in activation distributions leads to severe degradation in quantization accuracy for both activations and weights, significantly impairing inference performance; (2) Under limited memory, efficient offloading and collaborative inference of expert modules struggle to balance latency and throughput. To address these issues, this paper proposes an efficient MoE edge deployment scheme based on Hessian-Aware Quantization (HAQ) and CPU-GPU collaborative inference. First, by introducing smoothed Hessian matrix quantization, we achieve joint 8-bit quantization of activations and weights, which significantly alleviates the accuracy loss caused by outliers while ensuring efficient implementation on mainstream hardware. Second, we design an expert-level collaborative offloading and inference mechanism, which, combined with expert activation path statistics, enables efficient deployment and scheduling of expert modules between CPU and GPU, greatly reducing memory footprint and inference latency. Extensive experiments validate the effectiveness of our method on mainstream large models such as the OPT series and Mixtral 8*7B: on datasets like Wikitext2 and C4, the inference accuracy of the low-bit quantized model approaches that of the full-precision model, while GPU memory usage is reduced by about 60%, and inference latency is significantly improved.
Authors: Yuhao Liu, Rui Hu, Yu Chen, Longbo Huang
Abstract: Stochastic interpolants offer a robust framework for continuously transforming samples between arbitrary data distributions, holding significant promise for generative modeling. Despite their potential, rigorous finite-time convergence guarantees for practical numerical schemes remain largely unexplored. In this work, we address the finite-time convergence analysis of numerical implementations for ordinary differential equations (ODEs) derived from stochastic interpolants. Specifically, we establish novel finite-time error bounds in total variation distance for two widely used numerical integrators: the first-order forward Euler method and the second-order Heun's method. Furthermore, our analysis on the iteration complexity of specific stochastic interpolant constructions provides optimized schedules to enhance computational efficiency. Our theoretical findings are corroborated by numerical experiments, which validate the derived error bounds and complexity analyses.
Authors: Samiha Afaf Neha, Abir Ahammed Bhuiyan, Md. Ishrak Khan
Abstract: \textbf{Introduction:} Accurate prediction of Phage Virion Proteins (PVP) is essential for genomic studies due to their crucial role as structural elements in bacteriophages. Computational tools, particularly machine learning, have emerged for annotating phage protein sequences from high-throughput sequencing. However, effective annotation requires specialized sequence encodings. Our paper introduces ProteoKnight, a new image-based encoding method that addresses spatial constraints in existing techniques, yielding competitive performance in PVP classification using pre-trained convolutional neural networks. Additionally, our study evaluates prediction uncertainty in binary PVP classification through Monte Carlo Dropout (MCD). \textbf{Methods:} ProteoKnight adapts the classical DNA-Walk algorithm for protein sequences, incorporating pixel colors and adjusting walk distances to capture intricate protein features. Encoded sequences were classified using multiple pre-trained CNNs. Variance and entropy measures assessed prediction uncertainty across proteins of various classes and lengths. \textbf{Results:} Our experiments achieved 90.8% accuracy in binary classification, comparable to state-of-the-art methods. Multi-class classification accuracy remains suboptimal. Our uncertainty analysis unveils variability in prediction confidence influenced by protein class and sequence length. \textbf{Conclusions:} Our study surpasses frequency chaos game representation (FCGR) by introducing novel image encoding that mitigates spatial information loss limitations. Our classification technique yields accurate and robust PVP predictions while identifying low-confidence predictions.
Authors: Sibylle Marcotte, Gabriel Peyr\'e, R\'emi Gribonval
Abstract: A fundamental challenge in the theory of deep learning is to understand whether gradient-based training in high-dimensional parameter spaces can be captured by simpler, lower-dimensional structures, leading to so-called implicit bias. As a stepping stone, we study when a gradient flow on a high-dimensional variable $\theta$ implies an intrinsic gradient flow on a lower-dimensional variable $z = \phi(\theta)$, for an architecture-related function $\phi$. We express a so-called intrinsic dynamic property and show how it is related to the study of conservation laws associated with the factorization $\phi$. This leads to a simple criterion based on the inclusion of kernels of linear maps which yields a necessary condition for this property to hold. We then apply our theory to general ReLU networks of arbitrary depth and show that, for any initialization, it is possible to rewrite the flow as an intrinsic dynamic in a lower dimension that depends only on $z$ and the initialization, when $\phi$ is the so-called path-lifting. In the case of linear networks with $\phi$ the product of weight matrices, so-called balanced initializations are also known to enable such a dimensionality reduction; we generalize this result to a broader class of {\em relaxed balanced} initializations, showing that, in certain configurations, these are the \emph{only} initializations that ensure the intrinsic dynamic property. Finally, for the linear neural ODE associated with the limit of infinitely deep linear networks, with relaxed balanced initialization, we explicitly express the corresponding intrinsic dynamics.
Authors: Nikita Puchkin, Denis Suchkov, Alexey Naumov, Denis Belomestny
Abstract: Modern methods of generative modelling and unpaired image-to-image translation based on Schr\"odinger bridges and stochastic optimal control theory aim to transform an initial density to a target one in an optimal way. In the present paper, we assume that we only have access to i.i.d. samples from initial and final distributions. This makes our setup suitable for both generative modelling and unpaired image-to-image translation. Relying on the stochastic optimal control approach, we choose an Ornstein-Uhlenbeck process as the reference one and estimate the corresponding Schr\"odinger potential. Introducing a risk function as the Kullback-Leibler divergence between couplings, we derive tight bounds on generalization ability of an empirical risk minimizer in a class of Schr\"odinger potentials including Gaussian mixtures. Thanks to the mixing properties of the Ornstein-Uhlenbeck process, we almost achieve fast rates of convergence up to some logarithmic factors in favourable scenarios. We also illustrate performance of the suggested approach with numerical experiments.
Authors: Behnoush Khavari, Mehran Shakerinava, Jayesh Khullar, Jerry Huang, Fran\c{c}ois Rivest, Siamak Ravanbakhsh, Sarath Chandar
Abstract: Recent work has shown that LRNN models such as S4D, Mamba, and DeltaNet lack state-tracking capability due to either time-invariant transition matrices or restricted eigenvalue ranges. To address this, input-dependent transition matrices, particularly those that are complex or non-triangular, have been proposed to enhance SSM performance on such tasks. While existing theorems demonstrate that both input-independent and non-negative SSMs are incapable of solving simple state-tracking tasks, such as parity, regardless of depth, they do not explore whether combining these two types in a multilayer SSM could help. We investigate this question for efficient SSMs with diagonal transition matrices and show that such combinations still fail to solve parity. This implies that a recurrence layer must both be input-dependent and include negative eigenvalues. Our experiments support this conclusion by analyzing an SSM model that combines S4D and Mamba layers.
Authors: Mohamad Louai Shehab, Alperen Tercan, Necmiye Ozay
Abstract: In this paper, we consider the problem of recovering time-varying reward functions from either optimal policies or demonstrations coming from a max entropy reinforcement learning problem. This problem is highly ill-posed without additional assumptions on the underlying rewards. However, in many applications, the rewards are indeed parsimonious, and some prior information is available. We consider two such priors on the rewards: 1) rewards are mostly constant and they change infrequently, 2) rewards can be represented by a linear combination of a small number of feature functions. We first show that the reward identification problem with the former prior can be recast as a sparsification problem subject to linear constraints. Moreover, we give a polynomial-time algorithm that solves this sparsification problem exactly. Then, we show that identifying rewards representable with the minimum number of features can be recast as a rank minimization problem subject to linear constraints, for which convex relaxations of rank can be invoked. In both cases, these observations lead to efficient optimization-based reward identification algorithms. Several examples are given to demonstrate the accuracy of the recovered rewards as well as their generalizability.
Authors: Md Sultanul Arifin, Abu Nowshed Sakib, Yeasir Rayhan, Tanzima Hashem
Abstract: Lightning, a common feature of severe meteorological conditions, poses significant risks, from direct human injuries to substantial economic losses. These risks are further exacerbated by climate change. Early and accurate prediction of lightning would enable preventive measures to safeguard people, protect property, and minimize economic losses. In this paper, we present DeepLight, a novel deep learning architecture for predicting lightning occurrences. Existing prediction models face several critical limitations: they often struggle to capture the dynamic spatial context and inherent uncertainty of lightning events, underutilize key observational data, such as radar reflectivity and cloud properties, and rely heavily on Numerical Weather Prediction (NWP) systems, which are both computationally expensive and highly sensitive to parameter settings. To overcome these challenges, DeepLight leverages multi-source meteorological data, including radar reflectivity, cloud properties, and historical lightning occurrences through a dual-encoder architecture. By employing multi-branch convolution techniques, it dynamically captures spatial correlations across varying extents. Furthermore, its novel Hazy Loss function explicitly addresses the spatio-temporal uncertainty of lightning by penalizing deviations based on proximity to true events, enabling the model to better learn patterns amidst randomness. Extensive experiments show that DeepLight improves the Equitable Threat Score (ETS) by 18%-30% over state-of-the-art methods, establishing it as a robust solution for lightning prediction.
Authors: Zhipeng Chang, Zhenye Wen, Xiaofei Zhao
Abstract: Existing operator learning methods rely on supervised training with high-fidelity simulation data, introducing significant computational cost. In this work, we propose the deep Onsager operator learning (DOOL) method, a novel unsupervised framework for solving dissipative equations. Rooted in the Onsager variational principle (OVP), DOOL trains a deep operator network by directly minimizing the OVP-defined Rayleighian functional, requiring no labeled data, and then proceeds in time explicitly through conservation/change laws for the solution. Another key innovation here lies in the spatiotemporal decoupling strategy: the operator's trunk network processes spatial coordinates exclusively, thereby enhancing training efficiency, while integrated external time stepping enables temporal extrapolation. Numerical experiments on typical dissipative equations validate the effectiveness of the DOOL method, and systematic comparisons with supervised DeepONet and MIONet demonstrate its enhanced performance. Extensions are made to cover the second-order wave models with dissipation that do not directly follow OVP.
Authors: Fernando Martinez, Tao Li, Yingdong Lu, Juntao Chen
Abstract: Integrated, end-to-end learning of representations and policies remains a cornerstone of deep reinforcement learning (RL). However, to address the challenge of learning effective features from a sparse reward signal, recent trends have shifted towards adding complex auxiliary objectives or fully decoupling the two processes, often at the cost of increased design complexity. This work proposes an alternative to both decoupling and naive end-to-end learning, arguing that performance can be significantly improved by structuring the interaction between distinct perception and control networks with a principled, game-theoretic dynamic. We formalize this dynamic by introducing the Stackelberg Coupled Representation and Reinforcement Learning (SCORER) framework, which models the interaction between perception and control as a Stackelberg game. The perception network (leader) strategically learns features to benefit the control network (follower), whose own objective is to minimize its Bellman error. We approximate the game's equilibrium with a practical two-timescale algorithm. Applied to standard DQN variants on benchmark tasks, SCORER improves sample efficiency and final performance. Our results show that performance gains can be achieved through principled algorithmic design of the perception-control dynamic, without requiring complex auxiliary objectives or architectures.
Authors: Wei Qian, Chenxu Zhao, Yangyi Li, Wenqian Ye, Mengdi Huai
Abstract: Currently, various uncertainty quantification methods have been proposed to provide certainty and probability estimates for deep learning models' label predictions. Meanwhile, with the growing demand for the right to be forgotten, machine unlearning has been extensively studied as a means to remove the impact of requested sensitive data from a pre-trained model without retraining the model from scratch. However, the vulnerabilities of such generated predictive uncertainties with regard to dedicated malicious unlearning attacks remain unexplored. To bridge this gap, for the first time, we propose a new class of malicious unlearning attacks against predictive uncertainties, where the adversary aims to cause the desired manipulations of specific predictive uncertainty results. We also design novel optimization frameworks for our attacks and conduct extensive experiments, including black-box scenarios. Notably, our extensive experiments show that our attacks are more effective in manipulating predictive uncertainties than traditional attacks that focus on label misclassifications, and existing defenses against conventional attacks are ineffective against our attacks.
Authors: Tiantian Yang, Zhiqian Chen
Abstract: Integrating multi-omics data, such as DNA methylation, mRNA expression, and microRNA (miRNA) expression, offers a comprehensive view of the biological mechanisms underlying disease. However, the high dimensionality and complex interactions among omics layers present major challenges for predictive modeling. We propose Multi-Omics integration with Tree-generated Graph Neural Network (MOTGNN), a novel and interpretable framework for binary disease classification. MOTGNN employs eXtreme Gradient Boosting (XGBoost) to perform omics-specific supervised graph construction, followed by modality-specific Graph Neural Networks (GNNs) for hierarchical representation learning, and a deep feedforward network for cross-omics integration. On three real-world disease datasets, MOTGNN outperforms state-of-the-art baselines by 5-10% in accuracy, ROC-AUC, and F1-score, and remains robust to severe class imbalance (e.g., 87.2% vs. 33.4% F1 on imbalanced data). The model maintains computational efficiency through sparse graphs (2.1-2.8 edges per node) and provides built-in interpretability, revealing both top-ranked biomarkers and the relative contributions of each omics modality. These results highlight MOTGNN's potential to improve both predictive accuracy and interpretability in multi-omics disease modeling.
Authors: Zijian Liu
Abstract: In Online Convex Optimization (OCO), when the stochastic gradient has a finite variance, many algorithms provably work and guarantee a sublinear regret. However, limited results are known if the gradient estimate has a heavy tail, i.e., the stochastic gradient only admits a finite $\mathsf{p}$-th central moment for some $\mathsf{p}\in\left(1,2\right]$. Motivated by it, this work examines different old algorithms for OCO (e.g., Online Gradient Descent) in the more challenging heavy-tailed setting. Under the standard bounded domain assumption, we establish new regrets for these classical methods without any algorithmic modification. Remarkably, these regret bounds are fully optimal in all parameters (can be achieved even without knowing $\mathsf{p}$), suggesting that OCO with heavy tails can be solved effectively without any extra operation (e.g., gradient clipping). Our new results have several applications. A particularly interesting one is the first provable convergence result for nonsmooth nonconvex optimization under heavy-tailed noise without gradient clipping. Furthermore, we explore broader settings (e.g., smooth OCO) and extend our ideas to optimistic algorithms to handle different cases simultaneously.
Authors: Ricardo Matos, Luis Roque, Vitor Cerqueira
Abstract: Deep learning approaches are increasingly relevant for time series forecasting tasks. Methods such as N-BEATS, which is built on stacks of multilayer perceptrons (MLPs) blocks, have achieved state-of-the-art results on benchmark datasets and competitions. N-BEATS is also more interpretable relative to other deep learning approaches, as it decomposes forecasts into different time series components, such as trend and seasonality. In this work, we present N-BEATS-MOE, an extension of N-BEATS based on a Mixture-of-Experts (MoE) layer. N-BEATS-MOE employs a dynamic block weighting strategy based on a gating network which allows the model to better adapt to the characteristics of each time series. We also hypothesize that the gating mechanism provides additional interpretability by identifying which expert is most relevant for each series. We evaluate our method across 12 benchmark datasets against several approaches, achieving consistent improvements on several datasets, especially those composed of heterogeneous time series.
Authors: Yueyang Quan, Chang Wang, Shengjie Zhai, Minghong Fang, Zhuqing Liu
Abstract: Decentralized min-max optimization allows multi-agent systems to collaboratively solve global min-max optimization problems by facilitating the exchange of model updates among neighboring agents, eliminating the need for a central server. However, sharing model updates in such systems carry a risk of exposing sensitive data to inference attacks, raising significant privacy concerns. To mitigate these privacy risks, differential privacy (DP) has become a widely adopted technique for safeguarding individual data. Despite its advantages, implementing DP in decentralized min-max optimization poses challenges, as the added noise can hinder convergence, particularly in non-convex scenarios with complex agent interactions in min-max optimization problems. In this work, we propose an algorithm called DPMixSGD (Differential Private Minmax Hybrid Stochastic Gradient Descent), a novel privacy-preserving algorithm specifically designed for non-convex decentralized min-max optimization. Our method builds on the state-of-the-art STORM-based algorithm, one of the fastest decentralized min-max solutions. We rigorously prove that the noise added to local gradients does not significantly compromise convergence performance, and we provide theoretical bounds to ensure privacy guarantees. To validate our theoretical findings, we conduct extensive experiments across various tasks and models, demonstrating the effectiveness of our approach.
Authors: Sichen Zhao, Wei Shao, Jeffrey Chan, Ziqi Xu, Flora Salim
Abstract: As deep spatio-temporal neural networks are increasingly utilised in urban computing contexts, the deployment of such methods can have a direct impact on users of critical urban infrastructure, such as public transport, emergency services, and traffic management systems. While many spatio-temporal methods focus on improving accuracy, fairness has recently gained attention due to growing evidence that biased predictions in spatio-temporal applications can disproportionately disadvantage certain demographic or geographic groups, thereby reinforcing existing socioeconomic inequalities and undermining the ethical deployment of AI in public services. In this paper, we propose a novel framework, FairDRL-ST, based on disentangled representation learning, to address fairness concerns in spatio-temporal prediction, with a particular focus on mobility demand forecasting. By leveraging adversarial learning and disentangled representation learning, our framework learns to separate attributes that contain sensitive information. Unlike existing methods that enforce fairness through supervised learning, which may lead to overcompensation and degraded performance, our framework achieves fairness in an unsupervised manner with minimal performance loss. We apply our framework to real-world urban mobility datasets and demonstrate its ability to close fairness gaps while delivering competitive predictive performance compared to state-of-the-art fairness-aware methods.
Authors: Tasfiq E. Alam, Md Manjurul Ahsan, Shivakumar Raman
Abstract: Accurate and interpretable bearing fault classification is critical for ensuring the reliability of rotating machinery, particularly under variable operating conditions where domain shifts can significantly degrade model performance. This study proposes a physics-informed multimodal convolutional neural network (CNN) with a late fusion architecture, integrating vibration and motor current signals alongside a dedicated physics-based feature extraction branch. The model incorporates a novel physics-informed loss function that penalizes physically implausible predictions based on characteristic bearing fault frequencies - Ball Pass Frequency Outer (BPFO) and Ball Pass Frequency Inner (BPFI) - derived from bearing geometry and shaft speed. Comprehensive experiments on the Paderborn University dataset demonstrate that the proposed physics-informed approach consistently outperforms a non-physics-informed baseline, achieving higher accuracy, reduced false classifications, and improved robustness across multiple data splits. To address performance degradation under unseen operating conditions, three transfer learning (TL) strategies - Target-Specific Fine-Tuning (TSFT), Layer-Wise Adaptation Strategy (LAS), and Hybrid Feature Reuse (HFR) - are evaluated. Results show that LAS yields the best generalization, with additional performance gains when combined with physics-informed modeling. Validation on the KAIST bearing dataset confirms the framework's cross-dataset applicability, achieving up to 98 percent accuracy. Statistical hypothesis testing further verifies significant improvements (p < 0.01) in classification performance. The proposed framework demonstrates the potential of integrating domain knowledge with data-driven learning to achieve robust, interpretable, and generalizable fault diagnosis for real-world industrial applications.
Authors: Keyuan Zhang, Yin Sun, Bo Ji
Abstract: We consider a remote inference system with multiple modalities, where a multimodal machine learning (ML) model performs real-time inference using features collected from remote sensors. As sensor observations may change dynamically over time, fresh features are critical for inference tasks. However, timely delivering features from all modalities is often infeasible due to limited network resources. To this end, we study a two-modality scheduling problem to minimize the ML model's inference error, which is expressed as a penalty function of AoI for both modalities. We develop an index-based threshold policy and prove its optimality. Specifically, the scheduler switches modalities when the current modality's index function exceeds a threshold. We show that the two modalities share the same threshold, and both the index functions and the threshold can be computed efficiently. The optimality of our policy holds for (i) general AoI functions that are \emph{non-monotonic} and \emph{non-additive} and (ii) \emph{heterogeneous} transmission times. Numerical results show that our policy reduces inference error by up to 55% compared to round-robin and uniform random policies, which are oblivious to the AoI-based inference error function. Our results shed light on how to improve remote inference accuracy by optimizing task-oriented AoI functions.
Authors: Stephan Rabanser
Abstract: Machine learning (ML) systems are increasingly deployed in high-stakes domains where reliability is paramount. This thesis investigates how uncertainty estimation can enhance the safety and trustworthiness of ML, focusing on selective prediction -- where models abstain when confidence is low. We first show that a model's training trajectory contains rich uncertainty signals that can be exploited without altering its architecture or loss. By ensembling predictions from intermediate checkpoints, we propose a lightweight, post-hoc abstention method that works across tasks, avoids the cost of deep ensembles, and achieves state-of-the-art selective prediction performance. Crucially, this approach is fully compatible with differential privacy (DP), allowing us to study how privacy noise affects uncertainty quality. We find that while many methods degrade under DP, our trajectory-based approach remains robust, and we introduce a framework for isolating the privacy-uncertainty trade-off. Next, we then develop a finite-sample decomposition of the selective classification gap -- the deviation from the oracle accuracy-coverage curve -- identifying five interpretable error sources and clarifying which interventions can close the gap. This explains why calibration alone cannot fix ranking errors, motivating methods that improve uncertainty ordering. Finally, we show that uncertainty signals can be adversarially manipulated to hide errors or deny service while maintaining high accuracy, and we design defenses combining calibration audits with verifiable inference. Together, these contributions advance reliable ML by improving, evaluating, and safeguarding uncertainty estimation, enabling models that not only make accurate predictions -- but also know when to say "I do not know".
Authors: Xingwu Chen, Miao Lu, Beining Wu, Difan Zou
Abstract: Using more test-time computation during language model inference, such as generating more intermediate thoughts or sampling multiple candidate answers, has proven effective in significantly improving model performance. This paper takes an initial step toward bridging the gap between practical language model inference and theoretical transformer analysis by incorporating randomness and sampling. We focus on in-context linear regression with continuous/binary coefficients, where our framework simulates language model decoding through noise injection and binary coefficient sampling. Through this framework, we provide detailed analyses of widely adopted inference techniques. Supported by empirical results, our theoretical framework and analysis demonstrate the potential for offering new insights into understanding inference behaviors in real-world language models.
Authors: Nisha Chandramoorthy, Adriaan de Clercq
Abstract: A curious phenomenon observed in some dynamical generative models is the following: despite learning errors in the score function or the drift vector field, the generated samples appear to shift \emph{along} the support of the data distribution but not \emph{away} from it. In this work, we investigate this phenomenon of \emph{robustness of the support} by taking a dynamical systems approach on the generating stochastic/deterministic process. Our perturbation analysis of the probability flow reveals that infinitesimal learning errors cause the predicted density to be different from the target density only on the data manifold for a wide class of generative models. Further, what is the dynamical mechanism that leads to the robustness of the support? We show that the alignment of the top Lyapunov vectors (most sensitive infinitesimal perturbation directions) with the tangent spaces along the boundary of the data manifold leads to robustness and prove a sufficient condition on the dynamics of the generating process to achieve this alignment. Moreover, the alignment condition is efficient to compute and, in practice, for robust generative models, automatically leads to accurate estimates of the tangent bundle of the data manifold. Using a finite-time linear perturbation analysis on samples paths as well as probability flows, our work complements and extends existing works on obtaining theoretical guarantees for generative models from a stochastic analysis, statistical learning and uncertainty quantification points of view. Our results apply across different dynamical generative models, such as conditional flow-matching and score-based generative models, and for different target distributions that may or may not satisfy the manifold hypothesis.
Authors: Zhenpeng Su, Leiyu Pan, Xue Bai, Dening Liu, Guanting Dong, Jiaming Huang, Wenping Hu, Fuzheng Zhang, Kun Gai, Guorui Zhou
Abstract: We present Klear-Reasoner, a model with long reasoning capabilities that demonstrates careful deliberation during problem solving, achieving outstanding performance across multiple benchmarks. Although there are already many excellent works related to inference models in the current community, there are still many problems with reproducing high-performance inference models due to incomplete disclosure of training details. This report provides an in-depth analysis of the reasoning model, covering the entire post-training workflow from data preparation and long Chain-of-Thought supervised fine-tuning (long CoT SFT) to reinforcement learning (RL), along with detailed ablation studies for each experimental component. For SFT data, our experiments show that a small number of high-quality data sources are more effective than a large number of diverse data sources, and that difficult samples can achieve better results without accuracy filtering. In addition, we investigate two key issues with current clipping mechanisms in RL: Clipping suppresses critical exploration signals and ignores suboptimal trajectories. To address these challenges, we propose Gradient-Preserving clipping Policy Optimization (GPPO) that gently backpropagates gradients from clipped tokens. GPPO not only enhances the model's exploration capacity but also improves its efficiency in learning from negative samples. Klear-Reasoner exhibits exceptional reasoning abilities in mathematics and programming, scoring 90.5% on AIME 2024, 83.2% on AIME 2025, 66.0% on LiveCodeBench V5 and 58.1% on LiveCodeBench V6.
Authors: Advait Parulekar, Litu Rout, Karthikeyan Shanmugam, Sanjay Shakkottai
Abstract: We study the problem of posterior sampling in the context of score based generative models. We have a trained score network for a prior $p(x)$, a measurement model $p(y|x)$, and are tasked with sampling from the posterior $p(x|y)$. Prior work has shown this to be intractable in KL (in the worst case) under well-accepted computational hardness assumptions. Despite this, popular algorithms for tasks such as image super-resolution, stylization, and reconstruction enjoy empirical success. Rather than establishing distributional assumptions or restricted settings under which exact posterior sampling is tractable, we view this as a more general "tilting" problem of biasing a distribution towards a measurement. Under minimal assumptions, we show that one can tractably sample from a distribution that is simultaneously close to the posterior of a noised prior in KL divergence and the true posterior in Fisher divergence. Intuitively, this combination ensures that the resulting sample is consistent with both the measurement and the prior. To the best of our knowledge these are the first formal results for (approximate) posterior sampling in polynomial time.
Authors: Huiqi Deng, Hongbin Pei, Quanshi Zhang, Mengnan Du
Abstract: Attribution explanation is a typical approach for explaining deep neural networks (DNNs), inferring an importance or contribution score for each input variable to the final output. In recent years, numerous attribution methods have been developed to explain DNNs. However, a persistent concern remains unresolved, i.e., whether and which attribution methods faithfully reflect the actual contribution of input variables to the decision-making process. The faithfulness issue undermines the reliability and practical utility of attribution explanations. We argue that these concerns stem from three core challenges. First, difficulties arise in comparing attribution methods due to their unstructured heterogeneity, differences in heuristics, formulations, and implementations that lack a unified organization. Second, most methods lack solid theoretical underpinnings, with their rationales remaining absent, ambiguous, or unverified. Third, empirically evaluating faithfulness is challenging without ground truth. Recent theoretical advances provide a promising way to tackle these challenges, attracting increasing attention. We summarize these developments, with emphasis on three key directions: (i) Theoretical unification, which uncovers commonalities and differences among methods, enabling systematic comparisons; (ii) Theoretical rationale, clarifying the foundations of existing methods; (iii) Theoretical evaluation, rigorously proving whether methods satisfy faithfulness principles. Beyond a comprehensive review, we provide insights into how these studies help deepen theoretical understanding, inform method selection, and inspire new attribution methods. We conclude with a discussion of promising open problems for further work.
Authors: Guanqun Ma, David Lenz, Hanqi Guo, Tom Peterka, Bei Wang
Abstract: Implicit continuous models, such as functional models and implicit neural networks, are an increasingly popular method for replacing discrete data representations with continuous, high-order, and differentiable surrogates. These models offer new perspectives on the storage, transfer, and analysis of scientific data. In this paper, we introduce the first framework to directly extract complex topological features -- contours, Jacobi sets, and ridge-valley graphs -- from a type of continuous implicit model known as multivariate functional approximation (MFA). MFA replaces discrete data with continuous piecewise smooth functions. Given an MFA model as the input, our approach enables direct extraction of complex topological features from the model, without reverting to a discrete representation of the model. Our work is easily generalizable to any continuous implicit model that supports the queries of function values and high-order derivatives. Our work establishes the building blocks for performing topological data analysis and visualization on implicit continuous models.
Authors: Jia Zhang, Yao Liu, Chen-Xi Zhang, Yi Liu, Yi-Xuan Jin, Lan-Zhe Guo, Yu-Feng Li
Abstract: Aligning Large Language Models (LLMs) with diverse human values requires moving beyond a single holistic "better-than" preference criterion. While collecting fine-grained, aspect-specific preference data is more reliable and scalable, existing methods like Direct Preference Optimization (DPO) struggle with the severe noise and conflicts inherent in such aggregated datasets. In this paper, we tackle this challenge from a data-centric perspective. We first derive the Direct Multi-Preference Optimization (DMPO) objective, and uncover a key Preference Divergence (PD) term that quantifies inter-aspect preference conflicts. Instead of using this term for direct optimization, we leverage it to formulate a novel, theoretically-grounded data selection principle. Our principle advocates for selecting a subset of high-consensus data-identified by the most negative PD values-for efficient DPO training. We prove the optimality of this strategy by analyzing the loss bounds of the DMPO objective in the selection problem. To operationalize our approach, we introduce practical methods of PD term estimation and length bias mitigation, thereby proposing our PD selection method. Evaluation on the UltraFeedback dataset with three varying conflict levels shows that our simple yet effective strategy achieves over 10% relative improvement against both the standard holistic preference and a stronger oracle using aggregated preference signals, all while boosting training efficiency and obviating the need for intractable holistic preference annotating, unlocking the potential of robust LLM alignment via fine-grained preference signals.
Authors: Xiaoxue Yang, Jaeha Lee, Anna-Katharina Dick, Jasper Timm, Fei Xie, Diogo Cruz
Abstract: While defenses against single-turn jailbreak attacks on Large Language Models (LLMs) have improved significantly, multi-turn jailbreaks remain a persistent vulnerability, often achieving success rates exceeding 70% against models optimized for single-turn protection. This work presents an empirical analysis of automated multi-turn jailbreak attacks across state-of-the-art models including GPT-4, Claude, and Gemini variants, using the StrongREJECT benchmark. Our findings challenge the perceived sophistication of multi-turn attacks: when accounting for the attacker's ability to learn from how models refuse harmful requests, multi-turn jailbreaking approaches are approximately equivalent to simply resampling single-turn attacks multiple times. Moreover, attack success is correlated among similar models, making it easier to jailbreak newly released ones. Additionally, for reasoning models, we find surprisingly that higher reasoning effort often leads to higher attack success rates. Our results have important implications for AI safety evaluation and the design of jailbreak-resistant systems. We release the source code at https://github.com/diogo-cruz/multi_turn_simpler
Authors: Hyeon-Ju Jeon, Jeon-Ho Kang, In-Hyuk Kwon, O-Joun Lee
Abstract: This study aims to discover spatial correlations between Earth observations and atmospheric states to improve the forecasting accuracy of global atmospheric state estimation, which are usually conducted using conventional numerical weather prediction (NWP) systems and is the beginning of weather forecasting. NWP systems predict future atmospheric states at fixed locations, which are called NWP grid points, by analyzing previous atmospheric states and newly acquired Earth observations without fixed locations. Thus, surrounding meteorological context and the changing locations of the observations make spatial correlations between atmospheric states and observations over time. To handle complicated spatial correlations, which change dynamically, we employ spatiotemporal graph neural networks (STGNNs) with structure learning. However, structure learning has an inherent limitation that this can cause structural information loss and over-smoothing problem by generating excessive edges. To solve this problem, we regulate edge sampling by adaptively determining node degrees and considering the spatial distances between NWP grid points and observations. We validated the effectiveness of the proposed method by using real-world atmospheric state and observation data from East Asia. Even in areas with high atmospheric variability, the proposed method outperformed existing STGNN models with and without structure learning.
Authors: Ihor Stepanov, Mykhailo Shtopko, Dmytro Vodianytskyi, Oleksandr Lukashov, Alexander Yavorskyi, Mykyta Yaroshenko
Abstract: Classification is one of the most widespread tasks in AI applications, serving often as the first step in filtering, sorting, and categorizing data. Since modern AI systems must handle large volumes of input data and early pipeline stages can propagate errors downstream, achieving high efficiency and accuracy is critical. Moreover, classification requirements can change dynamically based on user needs, necessitating models with strong zero-shot capabilities. While generative LLMs have become mainstream for zero-shot classification due to their versatility, they suffer from inconsistent instruction following and computational inefficiency. Cross-encoders, commonly used as rerankers in RAG pipelines, face a different bottleneck: they must process text-label pairs sequentially, significantly reducing efficiency with large label sets. Embedding-based approaches offer good efficiency but struggle with complex scenarios involving logical and semantic constraints. We propose GLiClass, a novel method that adapts the GLiNER architecture for sequence classification tasks. Our approach achieves strong accuracy and efficiency comparable to embedding-based methods, while maintaining the flexibility needed for zero-shot and few-shot learning scenarios. Additionally, we adapted proximal policy optimization (PPO) for multi-label text classification, enabling training classifiers in data-sparse conditions or from human feedback.
Authors: Hyobin Park, Jinwook Jung, Minseok Seo, Hyunsoo Choi, Deukjae Cho, Sekil Park, Dong-Geol Choi
Abstract: With the increase in maritime traffic and the mandatory implementation of the Automatic Identification System (AIS), the importance and diversity of maritime traffic analysis tasks based on AIS data, such as vessel trajectory prediction, anomaly detection, and collision risk assessment, is rapidly growing. However, existing approaches tend to address these tasks individually, making it difficult to holistically consider complex maritime situations. To address this limitation, we propose a novel framework, AIS-LLM, which integrates time-series AIS data with a large language model (LLM). AIS-LLM consists of a Time-Series Encoder for processing AIS sequences, an LLM-based Prompt Encoder, a Cross-Modality Alignment Module for semantic alignment between time-series data and textual prompts, and an LLM-based Multi-Task Decoder. This architecture enables the simultaneous execution of three key tasks: trajectory prediction, anomaly detection, and risk assessment of vessel collisions within a single end-to-end system. Experimental results demonstrate that AIS-LLM outperforms existing methods across individual tasks, validating its effectiveness. Furthermore, by integratively analyzing task outputs to generate situation summaries and briefings, AIS-LLM presents the potential for more intelligent and efficient maritime traffic management.
Authors: Xutong Liu, Baran Atalar, Xiangxiang Dai, Jinhang Zuo, Siwei Wang, John C. S. Lui, Wei Chen, Carlee Joe-Wong
Abstract: Large Language Models (LLMs) are revolutionizing how users interact with information systems, yet their high inference cost poses serious scalability and sustainability challenges. Caching inference responses, allowing them to be retrieved without another forward pass through the LLM, has emerged as one possible solution. Traditional exact-match caching, however, overlooks the semantic similarity between queries, leading to unnecessary recomputation. Semantic caching addresses this by retrieving responses based on semantic similarity, but introduces a fundamentally different cache eviction problem: one must account for mismatch costs between incoming queries and cached responses. Moreover, key system parameters, such as query arrival probabilities and serving costs, are often unknown and must be learned over time. Existing semantic caching methods are largely ad-hoc, lacking theoretical foundations and unable to adapt to real-world uncertainty. In this paper, we present a principled, learning-based framework for semantic cache eviction under unknown query and cost distributions. We formulate both offline optimization and online learning variants of the problem, and develop provably efficient algorithms with state-of-the-art guarantees. We also evaluate our framework on a synthetic dataset, showing that our proposed algorithms perform matching or superior performance compared with baselines.
Authors: Chenchen Lin, Xuehe Wang
Abstract: Federated learning (FL) enables collaborative model training across decentralized clients without sharing local data, thereby enhancing privacy and facilitating collaboration among clients connected via social networks. However, these social connections introduce privacy externalities: a client's privacy loss depends not only on its privacy protection strategy but also on the privacy decisions of others, propagated through the network via multi-hop interactions. In this work, we propose a socially-aware privacy-preserving FL mechanism that systematically quantifies indirect privacy leakage through a multi-hop propagation model. We formulate the server-client interaction as a two-stage Stackelberg game, where the server, as the leader, optimizes incentive policies, and clients, as followers, strategically select their privacy budgets, which determine their privacy-preserving levels by controlling the magnitude of added noise. To mitigate information asymmetry in networked privacy estimation, we introduce a mean-field estimator to approximate the average external privacy risk. We theoretically prove the existence and convergence of the fixed point of the mean-field estimator and derive closed-form expressions for the Stackelberg Nash Equilibrium. Despite being designed from a client-centric incentive perspective, our mechanism achieves approximately-optimal social welfare, as revealed by Price of Anarchy (PoA) analysis. Experiments on diverse datasets demonstrate that our approach significantly improves client utilities and reduces server costs while maintaining model performance, outperforming both Social-Agnostic (SA) baselines and methods that account for social externalities.
Authors: Yooseok Lim, ByoungJun Jeon, Seong-A Park, Jisoo Lee, Sae Won Choi, Chang Wook Jeong, Ho-Geol Ryu, Hongyeol Lee, Hyun-Lim Yang
Abstract: Sepsis, a life-threatening inflammatory response to infection, causes organ dysfunction, making early detection and optimal management critical. Previous reinforcement learning (RL) approaches to sepsis management rely primarily on structured data, such as lab results or vital signs, and on a dearth of a comprehensive understanding of the patient's condition. In this work, we propose a Multimodal Offline REinforcement learning for Clinical notes Leveraged Enhanced stAte Representation (MORE-CLEAR) framework for sepsis control in intensive care units. MORE-CLEAR employs pre-trained large-scale language models (LLMs) to facilitate the extraction of rich semantic representations from clinical notes, preserving clinical context and improving patient state representation. Gated fusion and cross-modal attention allow dynamic weight adjustment in the context of time and the effective integration of multimodal data. Extensive cross-validation using two public (MIMIC-III and MIMIC-IV) and one private dataset demonstrates that MORE-CLEAR significantly improves estimated survival rate and policy performance compared to single-modal RL approaches. To our knowledge, this is the first to leverage LLM capabilities within a multimodal offline RL for better state representation in medical applications. This approach can potentially expedite the treatment and management of sepsis by enabling reinforcement learning models to propose enhanced actions based on a more comprehensive understanding of patient conditions.
Authors: Hao Liu, Chun Yang, Zhang xiaoxing, Xiaobin Zhu
Abstract: Time series forecasting plays a significant role in finance, energy, meteorology, and IoT applications. Recent studies have leveraged the generalization capabilities of large language models (LLMs) to adapt to time series forecasting, achieving promising performance. However, existing studies focus on token-level modal alignment, instead of bridging the intrinsic modality gap between linguistic knowledge structures and time series data patterns, greatly limiting the semantic representation. To address this issue, we propose a novel Semantic-Enhanced LLM (SE-LLM) that explores the inherent periodicity and anomalous characteristics of time series to embed into the semantic space to enhance the token embedding. This process enhances the interpretability of tokens for LLMs, thereby activating the potential of LLMs for temporal sequence analysis. Moreover, existing Transformer-based LLMs excel at capturing long-range dependencies but are weak at modeling short-term anomalies in time-series data. Hence, we propose a plugin module embedded within self-attention that models long-term and short-term dependencies to effectively adapt LLMs to time-series analysis. Our approach freezes the LLM and reduces the sequence dimensionality of tokens, greatly reducing computational consumption. Experiments demonstrate the superiority performance of our SE-LLM against the state-of-the-art (SOTA) methods.
Authors: Philipp Huber, David Li, Juan Pedro Guti\'errez Hermosillo Muriedas, Deifilia Kieckhefen, Markus G\"otz, Achim Streit, Charlotte Debus
Abstract: The increasing demand for computational resources of training neural networks leads to a concerning growth in energy consumption. While parallelization has enabled upscaling model and dataset sizes and accelerated training, its impact on energy consumption is often overlooked. To close this research gap, we conducted scaling experiments for data-parallel training of two models, ResNet50 and FourCastNet, and evaluated the impact of parallelization parameters, i.e., GPU count, global batch size, and local batch size, on predictive performance, training time, and energy consumption. We show that energy consumption scales approximately linearly with the consumed resources, i.e., GPU hours; however, the respective scaling factor differs substantially between distinct model trainings and hardware, and is systematically influenced by the number of samples and gradient updates per GPU hour. Our results shed light on the complex interplay of scaling up neural network training and can inform future developments towards more sustainable AI research.
Authors: Jingya Wang, Xin Deng, Wenjie Wei, Dehao Zhang, Shuai Wang, Qian Sun, Jieyuan Zhang, Hanwen Liu, Ning Xie, Malu Zhang
Abstract: Leveraging the event-driven paradigm, Spiking Neural Networks (SNNs) offer a promising approach for constructing energy-efficient Transformer architectures. Compared to directly trained Spiking Transformers, ANN-to-SNN conversion methods bypass the high training costs. However, existing methods still suffer from notable limitations, failing to effectively handle nonlinear operations in Transformer architectures and requiring additional fine-tuning processes for pre-trained ANNs. To address these issues, we propose a high-performance and training-free ANN-to-SNN conversion framework tailored for Transformer architectures. Specifically, we introduce a Multi-basis Exponential Decay (MBE) neuron, which employs an exponential decay strategy and multi-basis encoding method to efficiently approximate various nonlinear operations. It removes the requirement for weight modifications in pre-trained ANNs. Extensive experiments across diverse tasks (CV, NLU, NLG) and mainstream Transformer architectures (ViT, RoBERTa, GPT-2) demonstrate that our method achieves near-lossless conversion accuracy with significantly lower latency. This provides a promising pathway for the efficient and scalable deployment of Spiking Transformers in real-world applications.
Authors: Jinghan Yang, Jiayu Weng
Abstract: Deep neural networks can memorize corrupted labels, making data quality critical for model performance, yet real-world datasets are frequently compromised by both label noise and input noise. This paper proposes a mutual information-based framework for data selection under hybrid noise scenarios that quantifies statistical dependencies between inputs and labels. We compute each sample's pointwise contribution to the overall mutual information and find that lower contributions indicate noisy or mislabeled instances. Empirical validation on MNIST with different synthetic noise settings demonstrates that the method effectively filters low-quality samples. Under label corruption, training on high-MI samples improves classification accuracy by up to 15\% compared to random sampling. Furthermore, the method exhibits robustness to benign input modifications, preserving semantically valid data while filtering truly corrupted samples.
Authors: Pietro Talli, Federico Mason, Federico Chiariotti, Andrea Zanella
Abstract: In this work, we address the problem of training Reinforcement Learning (RL) agents over communication networks. The RL paradigm requires the agent to instantaneously perceive the state evolution to infer the effects of its actions on the environment. This is impossible if the agent receives state updates over lossy or delayed wireless systems and thus operates with partial and intermittent information. In recent years, numerous frameworks have been proposed to manage RL with imperfect feedback; however, they often offer specific solutions with a substantial computational burden. To address these limits, we propose a novel architecture, named Homomorphic Robust Remote Reinforcement Learning (HR3L), that enables the training of remote RL agents exchanging observations across a non-ideal wireless channel. HR3L considers two units: the transmitter, which encodes meaningful representations of the environment, and the receiver, which decodes these messages and performs actions to maximize a reward signal. Importantly, HR3L does not require the exchange of gradient information across the wireless channel, allowing for quicker training and a lower communication overhead than state-of-the-art solutions. Experimental results demonstrate that HR3L significantly outperforms baseline methods in terms of sample efficiency and adapts to different communication scenarios, including packet losses, delayed transmissions, and capacity limitations.
Authors: Jialu Zhou, Dianxi Shi, Shaowu Yang, Xinyu Wei, Mingyue Yang, Leqian Li, Mengzhu Wang, Chunping Qiu
Abstract: Multi-Domain Continual Learning (MDCL) acquires knowledge from sequential tasks with shifting class sets and distribution. Despite the Parameter-Efficient Fine-Tuning (PEFT) methods can adapt for this dual heterogeneity, they still suffer from catastrophic forgetting and forward forgetting. To address these challenges, we propose a Two-Level Routing Grouped Mixture-of-Experts (TRGE) method. Firstly, TRGE dynamically expands the pre-trained CLIP model, assigning specific expert group for each task to mitigate catastrophic forgetting. With the number of experts continually grows in this process, TRGE maintains the static experts count within the group and introduces the intra-group router to alleviate routing overfitting caused by the increasing routing complexity. Meanwhile, we design an inter-group routing policy based on task identifiers and task prototype distance, which dynamically selects relevant expert groups and combines their outputs to enhance inter-task collaboration. Secondly, to get the correct task identifiers, we leverage Multimodal Large Language Models (MLLMs) which own powerful multimodal comprehension capabilities to generate semantic task descriptions and recognize the correct task identifier. Finally, to mitigate forward forgetting, we dynamically fuse outputs for unseen samples from the frozen CLIP model and TRGE adapter based on training progress, leveraging both pre-trained and learned knowledge. Through extensive experiments across various settings, our method outperforms other advanced methods with fewer trainable parameters.
Authors: Fengdi Che
Abstract: Offline reinforcement learning (RL) aims to optimize the return given a fixed dataset of agent trajectories without additional interactions with the environment. While algorithm development has progressed rapidly, significant theoretical advances have also been made in understanding the fundamental challenges of offline RL. However, bridging these theoretical insights with practical algorithm design remains an ongoing challenge. In this survey, we explore key intuitions derived from theoretical work and their implications for offline RL algorithms. We begin by listing the conditions needed for the proofs, including function representation and data coverage assumptions. Function representation conditions tell us what to expect for generalization, and data coverage assumptions describe the quality requirement of the data. We then examine counterexamples, where offline RL is not solvable without an impractically large amount of data. These cases highlight what cannot be achieved for all algorithms and the inherent hardness of offline RL. Building on techniques to mitigate these challenges, we discuss the conditions that are sufficient for offline RL. These conditions are not merely assumptions for theoretical proofs, but they also reveal the limitations of these algorithms and remind us to search for novel solutions when the conditions cannot be satisfied.
Authors: Haowen Wang, Yun Yue, Zhiling Ye, Shuowen Zhang, Lei Fan, Jiaxin Liang, Jiadi Jiang, Cheng Wei, Jingyuan Deng, Xudong Han, Ji Li, Chunxiao Guo, Peng Wei, Jian Wang, Jinjie Gu
Abstract: Alignment methodologies have emerged as a critical pathway for enhancing language model alignment capabilities. While SFT (supervised fine-tuning) accelerates convergence through direct token-level loss intervention, its efficacy is constrained by offline policy trajectory. In contrast, RL(reinforcement learning) facilitates exploratory policy optimization, but suffers from low sample efficiency and stringent dependency on high-quality base models. To address these dual challenges, we propose GRAO (Group Relative Alignment Optimization), a unified framework that synergizes the respective strengths of SFT and RL through three key innovations: 1) A multi-sample generation strategy enabling comparative quality assessment via reward feedback; 2) A novel Group Direct Alignment Loss formulation leveraging intra-group relative advantage weighting; 3) Reference-aware parameter updates guided by pairwise preference dynamics. Our theoretical analysis establishes GRAO's convergence guarantees and sample efficiency advantages over conventional approaches. Comprehensive evaluations across complex human alignment tasks demonstrate GRAO's superior performance, achieving 57.70\%,17.65\% 7.95\% and 5.18\% relative improvements over SFT, DPO, PPO and GRPO baselines respectively. This work provides both a theoretically grounded alignment framework and empirical evidence for efficient capability evolution in language models.
Authors: Martin Rektoris, Milan Pape\v{z}, V\'aclav \v{S}m\'idl, Tom\'a\v{s} Pevn\'y
Abstract: Deep generative models (DGMs) for graphs achieve impressively high expressive power thanks to very efficient and scalable neural networks. However, these networks contain non-linearities that prevent analytical computation of many standard probabilistic inference queries, i.e., these DGMs are considered \emph{intractable}. While recently proposed Probabilistic Graph Circuits (PGCs) address this issue by enabling \emph{tractable} probabilistic inference, they operate on dense graph representations with $\mathcal{O}(n^2)$ complexity for graphs with $n$ nodes and \emph{$m$ edges}. To address this scalability issue, we introduce Sparse PGCs, a new class of tractable generative models that operate directly on sparse graph representation, reducing the complexity to $\mathcal{O}(n + m)$, which is particularly beneficial for $m \ll n^2$. In the context of de novo drug design, we empirically demonstrate that SPGCs retain exact inference capabilities, improve memory efficiency and inference speed, and match the performance of intractable DGMs in key metrics.
Authors: Qiang He, Setareh Maghsudi
Abstract: Large language models (LLMs) are increasingly deployed in real-world applications that require careful balancing of multiple, often conflicting, objectives, such as informativeness versus conciseness, or helpfulness versus creativity. However, current alignment methods, primarily based on RLHF, optimize LLMs toward a single reward function, resulting in rigid behavior that fails to capture the complexity and diversity of human preferences. This limitation hinders the adaptability of LLMs to practical scenarios, making multi-objective alignment (MOA) a critical yet underexplored area. To bridge this gap, we propose Pareto Multi-Objective Alignment (PAMA), a principled and computationally efficient algorithm designed explicitly for MOA in LLMs. In contrast to computationally prohibitive multi-objective optimization (MOO) methods, PAMA transforms multi-objective RLHF into a convex optimization with a closed-form solution, significantly enhancing scalability. Traditional MOO approaches suffer from prohibitive O(n^2*d) complexity, where d represents the number of model parameters, typically in the billions for LLMs, rendering direct optimization infeasible. PAMA reduces this complexity to O(n) where n is the number of objectives, enabling optimization to be completed within milliseconds. We provide theoretical guarantees that PAMA converges to a Pareto stationary point, where no objective can be improved without degrading at least one other. Extensive experiments across language models ranging from 125M to 7B parameters demonstrate PAMA's robust and effective MOA capabilities, aligning with its theoretical advantages. PAMA provides a highly efficient solution to the MOA problem that was previously considered intractable, offering a practical and theoretically grounded approach to aligning LLMs with diverse human values, paving the way for versatile and adaptable real-world AI deployments.
Authors: Rahul Khorana
Abstract: Recent advances in molecular representation learning have produced highly effective encodings of molecules for numerous cheminformatics and bioinformatics tasks. However, extracting general chemical insight while balancing predictive accuracy, interpretability, and computational efficiency remains a major challenge. In this work, we introduce a novel Graph Neural Network (GNN) architecture that combines compressed higher-order topological signals with standard molecular features. Our approach captures global geometric information while preserving computational tractability and human-interpretable structure. We evaluate our model across a range of benchmarks, from small-molecule datasets to complex material datasets, and demonstrate superior performance using a parameter-efficient architecture. We achieve the best performing results in both accuracy and robustness across almost all benchmarks. We open source all code \footnote{All code and results can be found on Github https://github.com/rahulkhorana/TFC-PACT-Net}.
Authors: Huanyu Liu, Jia Li, Chang Yu, Taozhi Chen, Yihong Dong, Lecheng Wang, Hu XiaoLong, Ge Li
Abstract: Reinforcement learning with verifiable reward (RLVR) has become a promising paradigm for post-training large language models (LLMs) to improve their reasoning capability. However, when the rollout accuracy is low on hard problems, the reward becomes sparse, limiting learning efficiency and causing exploration bottlenecks. Existing approaches either rely on stronger LLMs for distillation or filter out difficult problems, which limits scalability or restricts reasoning improvement through exploration. We propose EvoCoT, a self-evolving curriculum learning framework based on two-stage chain-of-thought (CoT) reasoning optimization. EvoCoT constrains the exploration space by self-generating and verifying CoT trajectories, then gradually shortens them to expand the space in a controlled way. This enables LLMs to stably learn from initially unsolved hard problems under sparse rewards. We apply EvoCoT to multiple LLM families, including Qwen, DeepSeek, and Llama. Experiments show that EvoCoT enables LLMs to solve previously unsolved problems, improves reasoning capability without external CoT supervision, and is compatible with various RL fine-tuning methods. We release the source code to support future research.
Authors: Carlo Cena, Mauro Martini, Marcello Chiaberge
Abstract: Attitude control is a fundamental aspect of spacecraft operations. Model Predictive Control (MPC) has emerged as a powerful strategy for these tasks, relying on accurate models of the system dynamics to optimize control actions over a prediction horizon. In scenarios where physics models are incomplete, difficult to derive, or computationally expensive, machine learning offers a flexible alternative by learning the system behavior directly from data. However, purely data-driven models often struggle with generalization and stability, especially when applied to inputs outside their training domain. To address these limitations, we investigate the benefits of incorporating Physics-Informed Neural Networks (PINNs) into the learning of spacecraft attitude dynamics, comparing their performance with that of purely data-driven approaches. Using a Real-valued Non-Volume Preserving (Real NVP) neural network architecture with a self-attention mechanism, we trained several models on simulated data generated with the Basilisk simulator. Two training strategies were considered: a purely data-driven baseline and a physics-informed variant to improve robustness and stability. Our results demonstrate that the inclusion of physics-based information significantly enhances the performance in terms of the mean relative error of the best architectures found by 27.08%. These advantages are particularly evident when the learned models are integrated into an MPC framework, where PINN-based models consistently outperform their purely data-driven counterparts in terms of control accuracy and robustness, yielding improvements of up to 42.86% in performance stability error and increased robustness-to-noise.
Authors: Sabrina Namazova, Alessandra Brondetta, Younes Strittmatter, Matthew Nassar, Sebastian Musslick
Abstract: Simulators have revolutionized scientific practice across the natural sciences. By generating data that reliably approximate real-world phenomena, they enable scientists to accelerate hypothesis testing and optimize experimental designs. This is perhaps best illustrated by AlphaFold, a Nobel-prize winning simulator in chemistry that predicts protein structures from amino acid sequences, enabling rapid prototyping of molecular interactions, drug targets, and protein functions. In the behavioral sciences, a reliable participant simulator - a system capable of producing human-like behavior across cognitive tasks - would represent a similarly transformative advance. Recently, Binz et al. introduced Centaur, a large language model (LLM) fine-tuned on human data from 160 experiments, proposing its use not only as a model of cognition but also as a participant simulator for "in silico prototyping of experimental studies", e.g., to advance automated cognitive science. Here, we review the core criteria for a participant simulator and assess how well Centaur meets them. Although Centaur demonstrates strong predictive accuracy, its generative behavior - a critical criterion for a participant simulator - systematically diverges from human data. This suggests that, while Centaur is a significant step toward predicting human behavior, it does not yet meet the standards of a reliable participant simulator or an accurate model of cognition.
Authors: Liang Hou, Yuan Gao, Boyuan Jiang, Xin Tao, Qi Yan, Renjie Liao, Pengfei Wan, Di Zhang, Kun Gai
Abstract: Diffusion models have achieved remarkable success in generative modeling. However, this study confirms the existence of overfitting in diffusion model training, particularly in data-limited regimes. To address this challenge, we propose Score Augmentation (ScoreAug), a novel data augmentation framework specifically designed for diffusion models. Unlike conventional augmentation approaches that operate on clean data, ScoreAug applies transformations to noisy data, aligning with the inherent denoising mechanism of diffusion. Crucially, ScoreAug further requires the denoiser to predict the augmentation of the original target. This design establishes an equivariant learning objective, enabling the denoiser to learn scores across varied denoising spaces, thereby realizing what we term score augmentation. We also theoretically analyze the relationship between scores in different spaces under general transformations. In experiments, we extensively validate ScoreAug on multiple benchmarks including CIFAR-10, FFHQ, AFHQv2, and ImageNet, with results demonstrating significant performance improvements over baselines. Notably, ScoreAug effectively mitigates overfitting across diverse scenarios, such as varying data scales and model capacities, while exhibiting stable convergence properties. Another advantage of ScoreAug over standard data augmentation lies in its ability to circumvent data leakage issues under certain conditions. Furthermore, we show that ScoreAug can be synergistically combined with traditional data augmentation techniques to achieve additional performance gains.
Authors: Amal Saadallah, Abdulaziz Al-Ademi
Abstract: Time series forecasting poses significant challenges in non-stationary environments where underlying patterns evolve over time. In this work, we propose a novel framework that enhances deep neural network (DNN) performance by leveraging specialized model adaptation and selection. Initially, a base DNN is trained offline on historical time series data. A reserved validation subset is then segmented to extract and cluster the most dominant patterns within the series, thereby identifying distinct regimes. For each identified cluster, the base DNN is fine-tuned to produce a specialized version that captures unique pattern characteristics. At inference, the most recent input is matched against the cluster centroids, and the corresponding fine-tuned version is deployed based on the closest similarity measure. Additionally, our approach integrates a concept drift detection mechanism to identify and adapt to emerging patterns caused by non-stationary behavior. The proposed framework is generalizable across various DNN architectures and has demonstrated significant performance gains on both traditional DNNs and recent advanced architectures implemented in the GluonTS library.
Authors: Richard J. Fawley, Renato Cordeiro de Amorim
Abstract: Clustering algorithms often assume all features contribute equally to the data structure, an assumption that usually fails in high-dimensional or noisy settings. Feature weighting methods can address this, but most require additional parameter tuning. We propose SHARK (Shapley Reweighted $k$-means), a feature-weighted clustering algorithm motivated by the use of Shapley values from cooperative game theory to quantify feature relevance, which requires no additional parameters beyond those in $k$-means. We prove that the $k$-means objective can be decomposed into a sum of per-feature Shapley values, providing an axiomatic foundation for unsupervised feature relevance and reducing Shapley computation from exponential to polynomial time. SHARK iteratively re-weights features by the inverse of their Shapley contribution, emphasising informative dimensions and down-weighting irrelevant ones. Experiments on synthetic and real-world data sets show that SHARK consistently matches or outperforms existing methods, achieving superior robustness and accuracy, particularly in scenarios where noise may be present. Software: https://github.com/rickfawley/shark.
Authors: Junyu Wu, Weiming Chang, Xiaotao Liu, Guanyou He, Tingfeng Xian, Haoqiang Hong, Boqi Chen, Haotao Tian, Tao Yang, Yunsheng Shi, Feng Lin, Ting Yao
Abstract: Reinforcement Learning from Human Feedback (RLHF) has emerged as a prominent paradigm for training large language models and multimodal systems. Despite notable advances enabled by existing RLHF training frameworks, significant challenges remain in scaling to complex multimodal workflows and adapting to dynamic workloads. In particular, current systems often encounter limitations related to controller scalability when managing large models, as well as inefficiencies in orchestrating intricate RLHF pipelines, especially in scenarios that require dynamic sampling and resource allocation. In this paper, we introduce WeChat-YATT (Yet Another Transformer Trainer in WeChat), a simple, scalable, and balanced RLHF training framework specifically designed to address these challenges. WeChat-YATT features a parallel controller programming model that enables flexible and efficient orchestration of complex RLHF workflows, effectively mitigating the bottlenecks associated with centralized controller architectures and facilitating scalability in large-scale data scenarios. In addition, we propose a dynamic placement schema that adaptively partitions computational resources and schedules workloads, thereby significantly reducing hardware idle time and improving GPU utilization under variable training conditions. We evaluate WeChat-YATT across a range of experimental scenarios, demonstrating that it achieves substantial improvements in throughput compared to state-of-the-art RLHF training frameworks. Furthermore, WeChat-YATT has been successfully deployed to train models supporting WeChat product features for a large-scale user base, underscoring its effectiveness and robustness in real-world applications.
Authors: Hongxin Yu, Yibing Wang, Fengyue Jin, Meng Zhang, Anni Chen
Abstract: Traffic state estimation (TSE) falls methodologically into three categories: model-driven, data-driven, and model-data dual-driven. Model-driven TSE relies on macroscopic traffic flow models originated from hydrodynamics. Data-driven TSE leverages historical sensing data and employs statistical models or machine learning methods to infer traffic state. Model-data dual-driven traffic state estimation attempts to harness the strengths of both aspects to achieve more accurate TSE. From the perspective of mathematical operator theory, TSE can be viewed as a type of operator that maps available measurements of inerested traffic state into unmeasured traffic state variables in real time. For the first time this paper proposes to study real-time freeway TSE in the idea of physics-informed deep operator network (PI-DeepONet), which is an operator-oriented architecture embedding traffic flow models based on deep neural networks. The paper has developed an extended architecture from the original PI-DeepONet. The extended architecture is featured with: (1) the acceptance of 2-D data input so as to support CNN-based computations; (2) the introduction of a nonlinear expansion layer, an attention mechanism, and a MIMO mechanism; (3) dedicated neural network design for adaptive identification of traffic flow model parameters. A traffic state estimator built on the basis of this extended PI-DeepONet architecture was evaluated with respect to a short freeway stretch of NGSIM and a large-scale urban expressway in China, along with other four baseline TSE methods. The evaluation results demonstrated that this novel TSE method outperformed the baseline methods with high-precision estimation results of flow and mean speed.
Authors: Xiang Li, Shanshan Wang, Chenglong Xiao
Abstract: Extensive experiments and prior studies show that no single maximum clique algorithm consistently performs best across all instances, highlighting the importance of selecting suitable algorithms based on instance features. Through an extensive analysis of relevant studies, it is found that there is a lack of research work concerning algorithm selection oriented toward the Maximum Clique Problem (MCP). In this work, we propose a learning-based framework that integrates both traditional machine learning and graph neural networks to address this gap. We construct a labeled dataset by running four exact MCP algorithms on a diverse collection of graph instances, accompanied by structural and global statistical features extracted from each graph. We first evaluate four conventional classifiers: Support Vector Machine (SVM), Random Forest (RF), Decision Tree (DT), and K-Nearest Neighbors (KNN), across multiple dataset variants. Experimental results show that RF consistently shows strong performance across metrics and dataset variants, making it a reliable baseline. In addition, feature importance analysis indicates that connectivity and topological structure are strong predictors of algorithm performance. Building on these findings, we develop a dual-channel model named GAT-MLP, which combines a Graph Attention Network (GAT) for local structural encoding with a Multilayer Perceptron (MLP) for global feature modeling. The GAT-MLP model shows strong and consistent performance across all metrics. Our results highlight the effectiveness of dual-channel architectures and the promise of graph neural networks in combinatorial algorithm selection.
Authors: Mohamad Assaad, Zeinab Nehme, Merouane Debbah
Abstract: Federated Learning (FL) is an emerging learning framework that enables edge devices to collaboratively train ML models without sharing their local data. FL faces, however, a significant challenge due to the high amount of information that must be exchanged between the devices and the aggregator in the training phase, which can exceed the limited capacity of wireless systems. In this paper, two communication-efficient FL methods are considered where communication overhead is reduced by communicating scalar values instead of long vectors and by allowing high number of users to send information simultaneously. The first approach employs a zero-order optimization technique with two-point gradient estimator, while the second involves a first-order gradient computation strategy. The novelty lies in leveraging channel information in the learning algorithms, eliminating hence the need for additional resources to acquire channel state information (CSI) and to remove its impact, as well as in considering asynchronous devices. We provide a rigorous analytical framework for the two methods, deriving convergence guarantees and establishing appropriate performance bounds.
Authors: Roksana Yahyaabadi, Ghazal Farhani, Taufiq Rahman, Soodeh Nikan, Abdullah Jirjees, Fadi Araji
Abstract: Accurate power consumption prediction is crucial for improving efficiency and reducing environmental impact, yet traditional methods relying on specialized instruments or rigid physical models are impractical for large-scale, real-world deployment. This study introduces a scalable data-driven method using powertrain dynamic feature sets and both traditional machine learning and deep neural networks to estimate instantaneous and cumulative power consumption in internal combustion engine (ICE), electric vehicle (EV), and hybrid electric vehicle (HEV) platforms. ICE models achieved high instantaneous accuracy with mean absolute error and root mean squared error on the order of $10^{-3}$, and cumulative errors under 3%. Transformer and long short-term memory models performed best for EVs and HEVs, with cumulative errors below 4.1% and 2.1%, respectively. Results confirm the approach's effectiveness across vehicles and models. Uncertainty analysis revealed greater variability in EV and HEV datasets than ICE, due to complex power management, emphasizing the need for robust models for advanced powertrains.
Authors: Maozhen Zhang, Mengnan Zhao, Bo Wang
Abstract: Prompt-based tuning has emerged as a lightweight alternative to full fine-tuning in large vision-language models, enabling efficient adaptation via learned contextual prompts. This paradigm has recently been extended to federated learning settings (e.g., PromptFL), where clients collaboratively train prompts under data privacy constraints. However, the security implications of prompt-based aggregation in federated multimodal learning remain largely unexplored, leaving a critical attack surface unaddressed. In this paper, we introduce \textbf{BadPromptFL}, the first backdoor attack targeting prompt-based federated learning in multimodal contrastive models. In BadPromptFL, compromised clients jointly optimize local backdoor triggers and prompt embeddings, injecting poisoned prompts into the global aggregation process. These prompts are then propagated to benign clients, enabling universal backdoor activation at inference without modifying model parameters. Leveraging the contextual learning behavior of CLIP-style architectures, BadPromptFL achieves high attack success rates (e.g., \(>90\%\)) with minimal visibility and limited client participation. Extensive experiments across multiple datasets and aggregation protocols validate the effectiveness, stealth, and generalizability of our attack, raising critical concerns about the robustness of prompt-based federated learning in real-world deployments.
Authors: Supriyo Chakraborty, Krishnan Raghavan
Abstract: The stability-plasticity dilemma, closely related to a neural network's (NN) capacity-its ability to represent tasks-is a fundamental challenge in continual learning (CL). Within this context, we introduce CL's effective model capacity (CLEMC) that characterizes the dynamic behavior of the stability-plasticity balance point. We develop a difference equation to model the evolution of the interplay between the NN, task data, and optimization procedure. We then leverage CLEMC to demonstrate that the effective capacity-and, by extension, the stability-plasticity balance point is inherently non-stationary. We show that regardless of the NN architecture or optimization method, a NN's ability to represent new tasks diminishes when incoming task distributions differ from previous ones. We conduct extensive experiments to support our theoretical findings, spanning a range of architectures-from small feedforward network and convolutional networks to medium-sized graph neural networks and transformer-based large language models with millions of parameters.
Authors: Sven Weinzierl, Sandra Zilker, Annina Liessmann, Martin K\"appel, Weixin Wang, Martin Matzner
Abstract: Event logs reflect the behavior of business processes that are mapped in organizational information systems. Predictive process monitoring (PPM) transforms these data into value by creating process-related predictions that provide the insights required for proactive interventions at process runtime. Existing PPM techniques require sufficient amounts of event data or other relevant resources that might not be readily available, preventing some organizations from utilizing PPM. The transfer learning-based PPM technique presented in this paper allows organizations without suitable event data or other relevant resources to implement PPM for effective decision support. The technique is instantiated in two real-life use cases, based on which numerical experiments are performed using event logs for IT service management processes in an intra- and inter-organizational setting. The results of the experiments suggest that knowledge of one business process can be transferred to a similar business process in the same or a different organization to enable effective PPM in the target context. With the proposed technique, organizations can benefit from transfer learning in an intra- and inter-organizational setting, where resources like pre-trained models are transferred within and across organizational boundaries.
Authors: Yunqing Li, Zixiang Tang, Jiaying Zhuang, Zhenyu Yang, Farhad Ameri, Jianbang Zhang
Abstract: Connecting an ever-expanding catalogue of products with suitable manufacturers and suppliers is critical for resilient, efficient global supply chains, yet traditional methods struggle to capture complex capabilities, certifications, geographic constraints, and rich multimodal data of real-world manufacturer profiles. To address these gaps, we introduce PMGraph, a public benchmark of bipartite and heterogeneous multimodal supply-chain graphs linking 8,888 manufacturers, over 70k products, more than 110k manufacturer-product edges, and over 29k product images. Building on this benchmark, we propose the Cascade Multimodal Attributed Graph C-MAG, a two-stage architecture that first aligns and aggregates textual and visual attributes into intermediate group embeddings, then propagates them through a manufacturer-product hetero-graph via multiscale message passing to enhance link prediction accuracy. C-MAG also provides practical guidelines for modality-aware fusion, preserving predictive performance in noisy, real-world settings.
Authors: Dimitris Tsaras, Xing Li, Lei Chen, Zhiyao Xie, Mingxuan Yuan
Abstract: In electronic design automation, logic optimization operators play a crucial role in minimizing the gate count of logic circuits. However, their computation demands are high. Operators such as refactor conventionally form iterative cuts for each node, striving for a more compact representation - a task which often fails 98% on average. Prior research has sought to mitigate computational cost through parallelization. In contrast, our approach leverages a classifier to prune unsuccessful cuts preemptively, thus eliminating unnecessary resynthesis operations. Experiments on the refactor operator using the EPFL benchmark suite and 10 large industrial designs demonstrate that this technique can speedup logic optimization by 3.9x on average compared with the state-of-the-art ABC implementation.
Authors: Cas Oude Hoekstra, Floris den Hengst
Abstract: Symbolic Regression (SR) is a well-established framework for generating interpretable or white-box predictive models. Although SR has been successfully applied to create interpretable estimates of the average of the outcome, it is currently not well understood how it can be used to estimate the relationship between variables at other points in the distribution of the target variable. Such estimates of e.g. the median or an extreme value provide a fuller picture of how predictive variables affect the outcome and are necessary in high-stakes, safety-critical application domains. This study introduces Symbolic Quantile Regression (SQR), an approach to predict conditional quantiles with SR. In an extensive evaluation, we find that SQR outperforms transparent models and performs comparably to a strong black-box baseline without compromising transparency. We also show how SQR can be used to explain differences in the target distribution by comparing models that predict extreme and central outcomes in an airline fuel usage case study. We conclude that SQR is suitable for predicting conditional quantiles and understanding interesting feature influences at varying quantiles.
Authors: Amir Ali Panahi, Daniel Luder, Billy Wu, Gregory Offer, Dirk Uwe Sauer, Weihan Li
Abstract: Reliable digital twins of lithium-ion batteries must achieve high physical fidelity with sub-millisecond speed. In this work, we benchmark three operator-learning surrogates for the Single Particle Model (SPM): Deep Operator Networks (DeepONets), Fourier Neural Operators (FNOs) and a newly proposed parameter-embedded Fourier Neural Operator (PE-FNO), which conditions each spectral layer on particle radius and solid-phase diffusivity. Models are trained on simulated trajectories spanning four current families (constant, triangular, pulse-train, and Gaussian-random-field) and a full range of State-of-Charge (SOC) (0 % to 100 %). DeepONet accurately replicates constant-current behaviour but struggles with more dynamic loads. The basic FNO maintains mesh invariance and keeps concentration errors below 1 %, with voltage mean-absolute errors under 1.7 mV across all load types. Introducing parameter embedding marginally increases error, but enables generalisation to varying radii and diffusivities. PE-FNO executes approximately 200 times faster than a 16-thread SPM solver. Consequently, PE-FNO's capabilities in inverse tasks are explored in a parameter estimation task with Bayesian optimisation, recovering anode and cathode diffusivities with 1.14 % and 8.4 % mean absolute percentage error, respectively, and 0.5918 percentage points higher error in comparison with classical methods. These results pave the way for neural operators to meet the accuracy, speed and parametric flexibility demands of real-time battery management, design-of-experiments and large-scale inference. PE-FNO outperforms conventional neural surrogates, offering a practical path towards high-speed and high-fidelity electrochemical digital twins.
Authors: Md. Wasiul Haque, Sagar Dasgupta, Mizanur Rahman
Abstract: Reliable indoor navigation remains a significant challenge in complex environments, particularly where external positioning signals and dedicated infrastructures are unavailable. This research presents Grid2Guide, a hybrid navigation framework that combines the A* search algorithm with a Small Language Model (SLM) to generate clear, human-readable route instructions. The framework first conducts a binary occupancy matrix from a given indoor map. Using this matrix, the A* algorithm computes the optimal path between origin and destination, producing concise textual navigation steps. These steps are then transformed into natural language instructions by the SLM, enhancing interpretability for end users. Experimental evaluations across various indoor scenarios demonstrate the method's effectiveness in producing accurate and timely navigation guidance. The results validate the proposed approach as a lightweight, infrastructure-free solution for real-time indoor navigation support.
Authors: Keyan Rahimi, Md. Wasiul Haque, Sagar Dasgupta, Mizanur Rahman
Abstract: Indoor navigation remains a complex challenge due to the absence of reliable GPS signals and the architectural intricacies of large enclosed environments. This study presents an indoor localization and navigation approach that integrates vision-based localization with large language model (LLM)-based navigation. The localization system utilizes a ResNet-50 convolutional neural network fine-tuned through a two-stage process to identify the user's position using smartphone camera input. To complement localization, the navigation module employs an LLM, guided by a carefully crafted system prompt, to interpret preprocessed floor plan images and generate step-by-step directions. Experimental evaluation was conducted in a realistic office corridor with repetitive features and limited visibility to test localization robustness. The model achieved high confidence and an accuracy of 96% across all tested waypoints, even under constrained viewing conditions and short-duration queries. Navigation tests using ChatGPT on real building floor maps yielded an average instruction accuracy of 75%, with observed limitations in zero-shot reasoning and inference time. This research demonstrates the potential for scalable, infrastructure-free indoor navigation using off-the-shelf cameras and publicly available floor plans, particularly in resource-constrained settings like hospitals, airports, and educational institutions.
Authors: Mingrong Lin, Ke Deng, Zhengyang Wu, Zetao Zheng, Jie Li
Abstract: Knowledge Tracing (KT) is committed to capturing students' knowledge mastery from their historical interactions. Simulating students' memory states is a promising approach to enhance both the performance and interpretability of knowledge tracing models. Memory consists of three fundamental processes: encoding, storage, and retrieval. Although forgetting primarily manifests during the storage stage, most existing studies rely on a single, undifferentiated forgetting mechanism, overlooking other memory processes as well as personalized forgetting patterns. To address this, this paper proposes memoryKT, a knowledge tracing model based on a novel temporal variational autoencoder. The model simulates memory dynamics through a three-stage process: (i) Learning the distribution of students' knowledge memory features, (ii) Reconstructing their exercise feedback, while (iii) Embedding a personalized forgetting module within the temporal workflow to dynamically modulate memory storage strength. This jointly models the complete encoding-storage-retrieval cycle, significantly enhancing the model's perception capability for individual differences. Extensive experiments on four public datasets demonstrate that our proposed approach significantly outperforms state-of-the-art baselines.
Authors: Guanghao Jin, Yuan Liang, Yihan Ma, Jingpei Wu, Guoyang Liu
Abstract: Large-scale models pre-trained on Electroencephalography (EEG) have shown promise in clinical applications such as neurological disorder detection. However, the practical deployment of EEG-based large-scale models faces critical challenges such as limited labeled EEG data and suboptimal performance in clinical scenarios. To address these issues, we propose NeuroDx-LM, a novel large-scale model specifically designed for detecting EEG-based neurological disorders. Our key contributions include (i) a Selective Temporal-Frequency Embedding mechanism that adaptively captures complex temporal and spectral patterns in EEG signals; and (ii) a Progressive Feature-Aware Training strategy that refines feature representation in a two-stage process. In the first stage, our model learns the fundamental discriminative features of EEG activities; in the second stage, the model further extracts more specialized fine-grained features for accurate diagnostic performance. We evaluated NeuroDx-LM on the CHB-MIT and Schizophrenia datasets, achieving state-of-the-art performance in EEG-based seizure and schizophrenia detection, respectively. These results demonstrate the great potential of EEG-based large-scale models to advance clinical applicability. Our code is available at https://github.com/LetItBe12345/NeuroDx-LM.
Authors: Hadi Khorsand, Vahid Pourahmadi
Abstract: In the active learning paradigm, using an oracle to label data has always been a complex and expensive task, and with the emersion of large unlabeled data pools, it would be highly beneficial If we could achieve better results without relying on an oracle. This research introduces OFAL, an oracle-free active learning scheme that utilizes neural network uncertainty. OFAL uses the model's own uncertainty to transform highly confident unlabeled samples into informative uncertain samples. First, we start with separating and quantifying different parts of uncertainty and introduce Monte Carlo Dropouts as an approximation of the Bayesian Neural Network model. Secondly, by adding a variational autoencoder, we go on to generate new uncertain samples by stepping toward the uncertain part of latent space starting from a confidence seed sample. By generating these new informative samples, we can perform active learning and enhance the model's accuracy. Lastly, we try to compare and integrate our method with other widely used active learning sampling methods.
Authors: Pravallika Abbineni, Saoud Aldowaish, Colin Liechty, Soroosh Noorzad, Ali Ghazizadeh, Morteza Fayazi
Abstract: Conducting a comprehensive literature review is crucial for advancing circuit design methodologies. However, the rapid influx of state-of-the-art research, inconsistent data representation, and the complexity of optimizing circuit design objectives make this task significantly challenging. In this paper, we propose MuaLLM, an open-source multimodal Large Language Model (LLM) agent for circuit design assistance that integrates a hybrid Retrieval-Augmented Generation (RAG) framework with an adaptive vector database of circuit design research papers. Unlike conventional LLMs, the MuaLLM agent employs a Reason + Act (ReAct) workflow for iterative reasoning, goal-setting, and multi-step information retrieval. It functions as a question-answering design assistant, capable of interpreting complex queries and providing reasoned responses grounded in circuit literature. Its multimodal capabilities enable processing of both textual and visual data, facilitating more efficient and comprehensive analysis. The system dynamically adapts using intelligent search tools, automated document retrieval from the internet, and real-time database updates. Unlike conventional approaches constrained by model context limits, MuaLLM decouples retrieval from inference, enabling scalable reasoning over arbitrarily large corpora. At the maximum context length supported by standard LLMs, MuaLLM remains up to 10x less costly and 1.6x faster while maintaining the same accuracy. This allows rapid, no-human-in-the-loop database generation, overcoming the bottleneck of simulation-based dataset creation for circuits. To evaluate MuaLLM, we introduce two custom datasets: RAG-250, targeting retrieval and citation performance, and Reasoning-100 (Reas-100), focused on multistep reasoning in circuit design. MuaLLM achieves 90.1% recall on RAG-250, and 86.8% accuracy on Reas-100.
Authors: Moses Openja, Paolo Arcaini, Foutse Khomh, Fuyuki Ishikawa
Abstract: Deep neural networks (DNNs) are being utilized in various aspects of our daily lives, including high-stakes decision-making applications that impact individuals. However, these systems reflect and amplify bias from the data used during training and testing, potentially resulting in biased behavior and inaccurate decisions. For instance, having different misclassification rates between white and black sub-populations. However, effectively and efficiently identifying and correcting biased behavior in DNNs is a challenge. This paper introduces FairFLRep, an automated fairness-aware fault localization and repair technique that identifies and corrects potentially bias-inducing neurons in DNN classifiers. FairFLRep focuses on adjusting neuron weights associated with sensitive attributes, such as race or gender, that contribute to unfair decisions. By analyzing the input-output relationships within the network, FairFLRep corrects neurons responsible for disparities in predictive quality parity. We evaluate FairFLRep on four image classification datasets using two DNN classifiers, and four tabular datasets with a DNN model. The results show that FairFLRep consistently outperforms existing methods in improving fairness while preserving accuracy. An ablation study confirms the importance of considering fairness during both fault localization and repair stages. Our findings also show that FairFLRep is more efficient than the baseline approaches in repairing the network.
Authors: Cem Ata Baykara, Saurav Raj Pandey, Ali Burak \"Unal, Harlin Lee, Mete Akg\"un
Abstract: Developing accurate and generalizable epileptic seizure prediction models from electroencephalography (EEG) data across multiple clinical sites is hindered by patient privacy regulations and significant data heterogeneity (non-IID characteristics). Federated Learning (FL) offers a privacy-preserving framework for collaborative training, but standard aggregation methods like Federated Averaging (FedAvg) can be biased by dominant datasets in heterogeneous settings. This paper investigates FL for seizure prediction using a single EEG channel across four diverse public datasets (Siena, CHB-MIT, Helsinki, NCH), representing distinct patient populations (adult, pediatric, neonate) and recording conditions. We implement privacy-preserving global normalization and propose a Random Subset Aggregation strategy, where each client trains on a fixed-size random subset of its data per round, ensuring equal contribution during aggregation. Our results show that locally trained models fail to generalize across sites, and standard weighted FedAvg yields highly skewed performance (e.g., 89.0% accuracy on CHB-MIT but only 50.8% on Helsinki and 50.6% on NCH). In contrast, Random Subset Aggregation significantly improves performance on under-represented clients (accuracy increases to 81.7% on Helsinki and 68.7% on NCH) and achieves a superior macro-average accuracy of 77.1% and pooled accuracy of 80.0% across all sites, demonstrating a more robust and fair global model. This work highlights the potential of balanced FL approaches for building effective and generalizable seizure prediction systems in realistic, heterogeneous multi-hospital environments while respecting data privacy.
Authors: Vincent Perreault, Katsumi Inoue, Richard Labib, Alain Hertz
Abstract: Traditional neural networks have an impressive classification performance, but what they learn cannot be inspected, verified or extracted. Neural Logic Networks on the other hand have an interpretable structure that enables them to learn a logical mechanism relating the inputs and outputs with AND and OR operations. We generalize these networks with NOT operations and biases that take into account unobserved data and develop a rigorous logical and probabilistic modeling in terms of concept combinations to motivate their use. We also propose a novel factorized IF-THEN rule structure for the model as well as a modified learning algorithm. Our method improves the state-of-the-art in Boolean networks discovery and is able to learn relevant, interpretable rules in tabular classification, notably on an example from the medical field where interpretability has tangible value.
Authors: Nicole Lai-Tan, Xiao Gu, Marios G. Philiastides, Fani Deligianni
Abstract: Personalised music-based interventions offer a powerful means of supporting motor rehabilitation by dynamically tailoring auditory stimuli to provide external timekeeping cues, modulate affective states, and stabilise gait patterns. Generalisable Brain-Computer Interfaces (BCIs) thus hold promise for adapting these interventions across individuals. However, inter-subject variability in EEG signals, further compounded by movement-induced artefacts and motor planning differences, hinders the generalisability of BCIs and results in lengthy calibration processes. We propose Individual Tangent Space Alignment (ITSA), a novel pre-alignment strategy incorporating subject-specific recentering, distribution matching, and supervised rotational alignment to enhance cross-subject generalisation. Our hybrid architecture fuses Regularised Common Spatial Patterns (RCSP) with Riemannian geometry in parallel and sequential configurations, improving class separability while maintaining the geometric structure of covariance matrices for robust statistical computation. Using leave-one-subject-out cross-validation, `ITSA' demonstrates significant performance improvements across subjects and conditions. The parallel fusion approach shows the greatest enhancement over its sequential counterpart, with robust performance maintained across varying data conditions and electrode configurations. The code will be made publicly available at the time of publication.
Authors: Zihe Liu, Jiashun Liu, Yancheng He, Weixun Wang, Jiaheng Liu, Ling Pan, Xinyu Hu, Shaopan Xiong, Ju Huang, Jian Hu, Shengyi Huang, Siran Yang, Jiamang Wang, Wenbo Su, Bo Zheng
Abstract: Reinforcement learning for LLM reasoning has rapidly emerged as a prominent research area, marked by a significant surge in related studies on both algorithmic innovations and practical applications. Despite this progress, several critical challenges remain, including the absence of standardized guidelines for employing RL techniques and a fragmented understanding of their underlying mechanisms. Additionally, inconsistent experimental settings, variations in training data, and differences in model initialization have led to conflicting conclusions, obscuring the key characteristics of these techniques and creating confusion among practitioners when selecting appropriate techniques. This paper systematically reviews widely adopted RL techniques through rigorous reproductions and isolated evaluations within a unified open-source framework. We analyze the internal mechanisms, applicable scenarios, and core principles of each technique through fine-grained experiments, including datasets of varying difficulty, model sizes, and architectures. Based on these insights, we present clear guidelines for selecting RL techniques tailored to specific setups, and provide a reliable roadmap for practitioners navigating the RL for the LLM domain. Finally, we reveal that a minimalist combination of two techniques can unlock the learning capability of critic-free policies using vanilla PPO loss. The results demonstrate that our simple combination consistently improves performance, surpassing strategies like GRPO and DAPO.
Authors: Tong Yang, Yu Huang, Yingbin Liang, Yuejie Chi
Abstract: Transformers have demonstrated remarkable capabilities in multi-step reasoning tasks. However, understandings of the underlying mechanisms by which they acquire these abilities through training remain limited, particularly from a theoretical standpoint. This work investigates how transformers learn to solve symbolic multi-step reasoning problems through chain-of-thought processes, focusing on path-finding in trees. We analyze two intertwined tasks: a backward reasoning task, where the model outputs a path from a goal node to the root, and a more complex forward reasoning task, where the model implements two-stage reasoning by first identifying the goal-to-root path and then reversing it to produce the root-to-goal path. Our theoretical analysis, grounded in the dynamics of gradient descent, shows that trained one-layer transformers can provably solve both tasks with generalization guarantees to unseen trees. In particular, our multi-phase training dynamics for forward reasoning elucidate how different attention heads learn to specialize and coordinate autonomously to solve the two subtasks in a single autoregressive path. These results provide a mechanistic explanation of how trained transformers can implement sequential algorithmic procedures. Moreover, they offer insights into the emergence of reasoning abilities, suggesting that when tasks are structured to take intermediate chain-of-thought steps, even shallow multi-head transformers can effectively solve problems that would otherwise require deeper architectures.
Authors: Kai Yao, Marc Juarez
Abstract: Generative models are increasingly adopted in high-stakes domains, yet current deployments offer no mechanisms to verify the origin of model outputs. We address this gap by extending model fingerprinting techniques beyond the traditional collaborative setting to one where the model provider may act adversarially. To our knowledge, this is the first work to evaluate fingerprinting for provenance attribution under such a threat model. The methods rely on a trusted verifier that extracts secret fingerprints from the model's output space, unknown to the provider, and trains a model to predict and verify them. Our empirical evaluation shows that our methods achieve near-zero FPR@95%TPR for instances of GAN and diffusion models, even when tested on small modifications to the original architecture and training data. Moreover, the methods remain robust against adversarial attacks that actively modify the outputs to bypass detection. Source codes are available at https://github.com/PSMLab/authprint.
Authors: Mohammed-Khalil Ghali, Cecil Pang, Oscar Molina, Carlos Gershenson-Garcia, Daehan Won
Abstract: Accurate forecasting of commodity price spikes is vital for countries with limited economic buffers, where sudden increases can strain national budgets, disrupt import-reliant sectors, and undermine food and energy security. This paper introduces a hybrid forecasting framework that combines historical commodity price data with semantic signals derived from global economic news, using an agentic generative AI pipeline. The architecture integrates dual-stream Long Short-Term Memory (LSTM) networks with attention mechanisms to fuse structured time-series inputs with semantically embedded, fact-checked news summaries collected from 1960 to 2023. The model is evaluated on a 64-year dataset comprising normalized commodity price series and temporally aligned news embeddings. Results show that the proposed approach achieves a mean AUC of 0.94 and an overall accuracy of 0.91 substantially outperforming traditional baselines such as logistic regression (AUC = 0.34), random forest (AUC = 0.57), and support vector machines (AUC = 0.47). Additional ablation studies reveal that the removal of attention or dimensionality reduction leads to moderate declines in performance, while eliminating the news component causes a steep drop in AUC to 0.46, underscoring the critical value of incorporating real-world context through unstructured text. These findings demonstrate that integrating agentic generative AI with deep learning can meaningfully improve early detection of commodity price shocks, offering a practical tool for economic planning and risk mitigation in volatile market environments while saving the very high costs of operating a full generative AI agents pipeline.
Authors: Chaeyeon Han, Seung Jae Lieu, Uijeong Hwang, Subhrajit Guhathakurta
Abstract: This study examines how indoor and outdoor aesthetics, streetscapes, and neighborhood features shape customer satisfaction at eating and dining establishments (EDEs) across different urban contexts, varying in car dependency, in Washington, DC. Using review photos and street view images, computer vision models quantified perceived safety and visual appeal. Ordinal logistic regression analyzed their effects on Yelp ratings. Findings reveal that both indoor and outdoor environments significantly impact EDE ratings, while streetscape quality's influence diminishes in car-dependent areas. The study highlights the need for context-sensitive planning that integrates indoor and outdoor factors to enhance customer experiences in diverse settings.
Authors: Lei Jiang, Fan Chen
Abstract: Neural scaling laws have driven the development of increasingly large language models (LLMs) by linking accuracy improvements to growth in parameter count, dataset size, and compute. However, these laws overlook the carbon emissions that scale exponentially with LLM size. This paper presents \textit{CarbonScaling}, an analytical framework that extends neural scaling laws to incorporate both operational and embodied carbon in LLM training. By integrating models for neural scaling, GPU hardware evolution, parallelism optimization, and carbon estimation, \textit{CarbonScaling} quantitatively connects model accuracy to carbon footprint. Results show that while a power-law relationship between accuracy and carbon holds, real-world inefficiencies significantly increase the scaling factor. Hardware technology scaling reduces carbon emissions for small to mid-sized models, but offers diminishing returns for extremely large LLMs due to communication overhead and underutilized GPUs. Training optimizations-especially aggressive critical batch size scaling-help alleviate this inefficiency. \textit{CarbonScaling} offers key insights for training more sustainable and carbon-efficient LLMs.
Authors: Jiayuan Wang, Q. M. Jonathan Wu, Katsuya Suto, Ning Zhang
Abstract: Autonomous driving systems rely on panoptic driving perception that requires both precision and real-time performance. In this work, we propose RMT-PPAD, a real-time, transformer-based multi-task model that jointly performs object detection, drivable area segmentation, and lane line segmentation. We introduce a lightweight module, a gate control with an adapter to adaptively fuse shared and task-specific features, effectively alleviating negative transfer between tasks. Additionally, we design an adaptive segmentation decoder to learn the weights over multi-scale features automatically during the training stage. This avoids the manual design of task-specific structures for different segmentation tasks. We also identify and resolve the inconsistency between training and testing labels in lane line segmentation. This allows fairer evaluation. Experiments on the BDD100K dataset demonstrate that RMT-PPAD achieves state-of-the-art results with mAP50 of 84.9% and Recall of 95.4% for object detection, mIoU of 92.6% for drivable area segmentation, and IoU of 56.8% and accuracy of 84.7% for lane line segmentation. The inference speed reaches 32.6 FPS. Moreover, we introduce real-world scenarios to evaluate RMT-PPAD performance in practice. The results show that RMT-PPAD consistently delivers stable performance. The source codes and pre-trained models are released at https://github.com/JiayuanWang-JW/RMT-PPAD.
Authors: Ming-Kun Xie, Jia-Hao Xiao, Gang Niu, Lei Feng, Zhiqiang Kou, Min-Ling Zhang, Masashi Sugiyama
Abstract: Large Vision-Language Models (LVLMs), empowered by the success of Large Language Models (LLMs), have achieved impressive performance across domains. Despite the great advances in LVLMs, they still suffer from the unavailable object hallucination issue, which tends to generate objects inconsistent with the image content. The most commonly used Polling-based Object Probing Evaluation (POPE) benchmark evaluates this issue by sampling negative categories according to category-level statistics, \textit{e.g.}, category frequencies and co-occurrence. However, with the continuous advancement of LVLMs, the POPE benchmark has shown diminishing effectiveness in assessing object hallucination, as it employs a simplistic sampling strategy that overlooks image-specific information and restricts distractors to negative object categories only. In this paper, we introduce the Hallucination searching-based Object Probing Evaluation (HOPE) benchmark, aiming to generate the most misleading distractors (\textit{i.e.}, non-existent objects or incorrect image descriptions) that can trigger hallucination in LVLMs, which serves as a means to more rigorously assess their immunity to hallucination. To explore the image-specific information, the content-aware hallucination searching leverages Contrastive Language-Image Pre-Training (CLIP) to approximate the predictive behavior of LVLMs by selecting negative objects with the highest predicted likelihood as distractors. To expand the scope of hallucination assessment, the description-based hallucination searching constructs highly misleading distractors by pairing true objects with false descriptions. Experimental results show that HOPE leads to a precision drop of at least 9\% and up to 23\% across various state-of-the-art LVLMs, significantly outperforming POPE in exposing hallucination vulnerabilities. The code is available at https://github.com/xiemk/HOPE.
Authors: Faisal Ahmed
Abstract: Accurate classification of Acute Lymphoblastic Leukemia (ALL) from peripheral blood smear images is essential for early diagnosis and effective treatment planning. This study investigates the use of transfer learning with pretrained convolutional neural networks (CNNs) to improve diagnostic performance. To address the class imbalance in the dataset of 3,631 Hematologic and 7,644 ALL images, we applied extensive data augmentation techniques to create a balanced training set of 10,000 images per class. We evaluated several models, including ResNet50, ResNet101, and EfficientNet variants B0, B1, and B3. EfficientNet-B3 achieved the best results, with an F1-score of 94.30%, accuracy of 92.02%, andAUCof94.79%,outperformingpreviouslyreported methods in the C-NMCChallenge. Thesefindings demonstrate the effectiveness of combining data augmentation with advanced transfer learning models, particularly EfficientNet-B3, in developing accurate and robust diagnostic tools for hematologic malignancy detection.
Authors: Zhanye Luo, Yuefeng Han, Xiufan Yu
Abstract: Large language models (LLMs) generate text embeddings from text data, producing vector representations that capture the semantic meaning and contextual relationships of words. However, the high dimensionality of these embeddings often impedes efficiency and drives up computational cost in downstream tasks. To address this, we propose AutoEncoder-Augmented Learning with Text (AEALT), a supervised, factor-augmented framework that incorporates dimension reduction directly into pre-trained LLM workflows. First, we extract embeddings from text documents; next, we pass them through a supervised augmented autoencoder to learn low-dimensional, task-relevant latent factors. By modeling the nonlinear structure of complex embeddings, AEALT outperforms conventional deep-learning approaches that rely on raw embeddings. We validate its broad applicability with extensive experiments on classification, anomaly detection, and prediction tasks using multiple real-world public datasets. Numerical results demonstrate that AEALT yields substantial gains over both vanilla embeddings and several standard dimension reduction methods.
Authors: Yinqiu Huang, Hao Ma, Wenshuai Chen, Shuli Wang, Yongqiang Zhang, Xue Wei, Yinhua Zhu, Haitao Wang, Xingxing Wang
Abstract: Bid shading plays a crucial role in Real-Time Bidding~(RTB) by adaptively adjusting the bid to avoid advertisers overspending. Existing mainstream two-stage methods, which first model bid landscapes and then optimize surplus using operations research techniques, are constrained by unimodal assumptions that fail to adapt for non-convex surplus curves and are vulnerable to cascading errors in sequential workflows. Additionally, existing discretization models of continuous values ignore the dependence between discrete intervals, reducing the model's error correction ability, while sample selection bias in bidding scenarios presents further challenges for prediction. To address these issues, this paper introduces Generative Bid Shading~(GBS), which comprises two primary components: (1) an end-to-end generative model that utilizes an autoregressive approach to generate shading ratios by stepwise residuals, capturing complex value dependencies without relying on predefined priors; and (2) a reward preference alignment system, which incorporates a channel-aware hierarchical dynamic network~(CHNet) as the reward model to extract fine-grained features, along with modules for surplus optimization and exploration utility reward alignment, ultimately optimizing both short-term and long-term surplus using group relative policy optimization~(GRPO). Extensive experiments on both offline and online A/B tests validate GBS's effectiveness. Moreover, GBS has been deployed on the Meituan DSP platform, serving billions of bid requests daily.
Authors: Unisha Joshi
Abstract: The challenges associated with deepfake detection are increasing significantly with the latest advancements in technology and the growing popularity of deepfake videos and images. Despite the presence of numerous detection models, demographic bias in the deepfake dataset remains largely unaddressed. This paper focuses on the mitigation of age-specific bias in the deepfake dataset by introducing an age-diverse deepfake dataset that will improve fairness across age groups. The dataset is constructed through a modular pipeline incorporating the existing deepfake datasets Celeb-DF, FaceForensics++, and UTKFace datasets, and the creation of synthetic data to fill the age distribution gaps. The effectiveness and generalizability of this dataset are evaluated using three deepfake detection models: XceptionNet, EfficientNet, and LipForensics. Evaluation metrics, including AUC, pAUC, and EER, revealed that models trained on the age-diverse dataset demonstrated fairer performance across age groups, improved overall accuracy, and higher generalization across datasets. This study contributes a reproducible, fairness-aware deepfake dataset and model pipeline that can serve as a foundation for future research in fairer deepfake detection. The complete dataset and implementation code are available at https://github.com/unishajoshi/age-diverse-deepfake-detection.
URLs: https://github.com/unishajoshi/age-diverse-deepfake-detection.
Authors: Sarina Penquitt, Jonathan Klees, Rinor Cakaj, Daniel Kondermann, Matthias Rottmann, Lars Schmarje
Abstract: Object detection has advanced rapidly in recent years, driven by increasingly large and diverse datasets. However, label errors, defined as missing labels, incorrect classification or inaccurate localization, often compromise the quality of these datasets. This can have a significant impact on the outcomes of training and benchmark evaluations. Although several methods now exist for detecting label errors in object detection datasets, they are typically validated only on synthetic benchmarks or limited manual inspection. How to correct such errors systemically and at scale therefore remains an open problem. We introduce a semi-automated framework for label-error correction called REC$\checkmark$D (Rechecked). Building on existing detectors, the framework pairs their error proposals with lightweight, crowd-sourced microtasks. These tasks enable multiple annotators to independently verify each candidate bounding box, and their responses are aggregated to estimate ambiguity and improve label quality. To demonstrate the effectiveness of REC$\checkmark$D, we apply it to the class pedestrian in the KITTI dataset. Our crowdsourced review yields high-quality corrected annotations, which indicate a rate of at least 24% of missing and inaccurate annotations in original annotations. This validated set will be released as a new real-world benchmark for label error detection and correction. We show that current label error detection methods, when combined with our correction framework, can recover hundreds of errors in the time it would take a human to annotate bounding boxes from scratch. However, even the best methods still miss up to 66% of the true errors and with low quality labels introduce more errors than they find. This highlights the urgent need for further research, now enabled by our released benchmark.
Authors: Zihao Hu (The Chinese University of Hong Kong), Jia Yan (The Hong Kong University of Science and Technology), Ying-Jun Angela Zhang (The Chinese University of Hong Kong)
Abstract: The ever-growing learning model size nowadays challenges the communication efficiency and privacy preservation of the traditional federated learning (FL). In this paper, we propose a novel differentially private (DP) over-the-air federated distillation (FD) framework, where wireless devices (WDs) periodically share noise-perturbed model outputs with the parameter server by harnessing the superposition property of multi-access channels. Accordingly, over-the-air FD enables the shared responsibility of the DP preservation on the low-dimensional disclosed signals among WDs. We study the communication-learning co-design problem in differentially private over-the-air FD, aiming to maximize the learning convergence rate while meeting the transmit power and DP requirements of WDs. The main challenge is rooted in the intractable learning and privacy analysis in over-the-air FD, together with the strong coupling among the decision variables spanning two timescales. To tackle this problem, we first derive the analytical learning convergence rate and privacy losses of WDs, based on which the optimal transceiver design per FD round and long-term training rounds decision are obtained in the closed forms. Numerical results demonstrate that the proposed differentially private over-the-air FD approach achieves a better learning-privacy trade-off with largely-reduced communication overhead than the conventional FL benchmarks.
Authors: Simon Baur, Alexandra Benova, Emilio Dolgener Cant\'u, Jackie Ma
Abstract: Deploying deep learning models in clinical practice often requires leveraging multiple data modalities, such as images, text, and structured data, to achieve robust and trustworthy decisions. However, not all modalities are always available at inference time. In this work, we propose multimodal privileged knowledge distillation (MMPKD), a training strategy that utilizes additional modalities available solely during training to guide a unimodal vision model. Specifically, we used a text-based teacher model for chest radiographs (MIMIC-CXR) and a tabular metadata-based teacher model for mammography (CBIS-DDSM) to distill knowledge into a vision transformer student model. We show that MMPKD can improve the resulting attention maps' zero-shot capabilities of localizing ROI in input images, while this effect does not generalize across domains, as contrarily suggested by prior research.
Authors: Sina Baghal
Abstract: Pasur is a fishing card game played over six rounds and is played similarly to games such as Cassino and Scopa, and Bastra. This paper introduces a CUDA-accelerated computational framework for simulating Pasur, emphasizing efficient memory management. We use our framework to compute near-Nash equilibria via Counterfactual Regret Minimization (CFR), a well-known algorithm for solving large imperfect-information games. Solving Pasur presents unique challenges due to its intricate rules and the large size of its game tree. We handle rule complexity using PyTorch CUDA tensors and to address the memory-intensive nature of the game, we decompose the game tree into two key components: (1) actual game states, and (2) inherited scores from previous rounds. We construct the Full Game Tree by pairing card states with accumulated scores in the Unfolding Process. This design reduces memory overhead by storing only essential strategy values and node connections. To further manage computational complexity, we apply a round-by-round backward training strategy, starting from the final round and recursively propagating average utilities to earlier stages. Our approach constructs the complete game tree, which on average consists of over $10^9$ nodes. We provide detailed implementation snippets. After computing a near-Nash equilibrium strategy, we train a tree-based model to predict these strategies for use during gameplay. We then estimate the fair value of each deck through large-scale self-play between equilibrium strategies by simulating, for instance, 10,000 games per matchup, executed in parallel using GPU acceleration. Similar frameworks can be extended to other reinforcement learning algorithms where the action tree naturally decomposes into multiple rounds such as turn-based strategy games or sequential trading decisions in financial markets.
Authors: Jing Zhang, Xiaowei Yu, Minheng Chen, Lu Zhang, Tong Chen, Yan Zhuang, Chao Cao, Yanjun Lyu, Li Su, Tianming Liu, Dajiang Zhu
Abstract: Integrating brain imaging data with clinical reports offers a valuable opportunity to leverage complementary multimodal information for more effective and timely diagnosis in practical clinical settings. This approach has gained significant attention in brain disorder research, yet a key challenge remains: how to effectively link objective imaging data with subjective text-based reports, such as doctors' notes. In this work, we propose a novel framework that aligns brain connectomes with clinical reports in a shared cross-modal latent space at both the subject and connectome levels, thereby enhancing representation learning. The key innovation of our approach is that we treat brain subnetworks as tokens of imaging data, rather than raw image patches, to align with word tokens in clinical reports. This enables a more efficient identification of system-level associations between neuroimaging findings and clinical observations, which is critical since brain disorders often manifest as network-level abnormalities rather than isolated regional alterations. We applied our method to mild cognitive impairment (MCI) using the ADNI dataset. Our approach not only achieves state-of-the-art predictive performance but also identifies clinically meaningful connectome-text pairs, offering new insights into the early mechanisms of Alzheimer's disease and supporting the development of clinically useful multimodal biomarkers.
Authors: Mohammad Zia Ur Rehman, Anukriti Bhatnagar, Omkar Kabde, Shubhi Bansal, Nagendra Kumar
Abstract: The existing research has primarily focused on text and image-based hate speech detection, video-based approaches remain underexplored. In this work, we introduce a novel dataset, ImpliHateVid, specifically curated for implicit hate speech detection in videos. ImpliHateVid consists of 2,009 videos comprising 509 implicit hate videos, 500 explicit hate videos, and 1,000 non-hate videos, making it one of the first large-scale video datasets dedicated to implicit hate detection. We also propose a novel two-stage contrastive learning framework for hate speech detection in videos. In the first stage, we train modality-specific encoders for audio, text, and image using contrastive loss by concatenating features from the three encoders. In the second stage, we train cross-encoders using contrastive learning to refine multimodal representations. Additionally, we incorporate sentiment, emotion, and caption-based features to enhance implicit hate detection. We evaluate our method on two datasets, ImpliHateVid for implicit hate speech detection and another dataset for general hate speech detection in videos, HateMM dataset, demonstrating the effectiveness of the proposed multimodal contrastive learning for hateful content detection in videos and the significance of our dataset.
Authors: Xiaoyuan Zhu, Muru Zhang, Ollie Liu, Robin Jia, Willie Neiswanger
Abstract: Modern large language models often encode sensitive, harmful, or copyrighted knowledge, raising the need for post-hoc unlearning-the ability to remove specific domains of knowledge from a model without full retraining. A major bottleneck in current unlearning pipelines is constructing effective forget sets-datasets that approximate the target domain and guide the model to forget it. In this work, we introduce a scalable, automated approach to generate high-quality forget sets using language models themselves. Our method synthesizes textbook-style data through a structured prompting pipeline, requiring only a domain name as input. Through experiments on unlearning biosecurity, cybersecurity, and Harry Potter novels, we show that our synthetic datasets consistently outperform the baseline synthetic alternatives and are comparable to the expert-curated ones. Additionally, ablation studies reveal that the multi-step generation pipeline significantly boosts data diversity, which in turn improves unlearning utility. Overall, our findings suggest that synthetic datasets offer a promising path toward practical, scalable unlearning for a wide range of emerging domains without the need for manual intervention. We release our code and dataset at https://github.com/xyzhu123/Synthetic_Textbook.
Authors: Adedire D. Adesiji, Jiashuo Wang, Cheng-Shu Kuo, Keith A. Brown
Abstract: A key goal of modern materials science is accelerating the pace of materials discovery. Self-driving labs, or systems that select experiments using machine learning and then execute them using automation, are designed to fulfil this promise by performing experiments faster, more intelligently, more reliably, and with richer metadata than conventional means. This review summarizes progress in understanding the degree to which SDLs accelerate learning by quantifying how much they reduce the number of experiments required for a given goal. The review begins by summarizing the theory underlying two key metrics, namely acceleration factor AF and enhancement factor EF, which quantify how much faster and better an algorithm is relative to a reference strategy. Next, we provide a comprehensive review of the literature, which reveals a wide range of AFs with a median of 6, and that tends to increase with the dimensionality of the space, reflecting an interesting blessing of dimensionality. In contrast, reported EF values vary by over two orders of magnitude, although they consistently peak at 10-20 experiments per dimension. To understand these results, we perform a series of simulated Bayesian optimization campaigns that reveal how EF depends upon the statistical properties of the parameter space while AF depends on its complexity. Collectively, these results reinforce the motivation for using SDLs by revealing their value across a wide range of material parameter spaces and provide a common language for quantifying and understanding this acceleration.
Authors: Jingmao Li, Yuanxing Chen, Shuangge Ma, Kuangnan Fang
Abstract: Federated learning has emerged as an essential paradigm for distributed multi-source data analysis under privacy concerns. Most existing federated learning methods focus on the ``static" datasets. However, in many real-world applications, data arrive continuously over time, forming streaming datasets. This introduces additional challenges for data storage and algorithm design, particularly under high-dimensional settings. In this paper, we propose a federated online learning (FOL) method for distributed multi-source streaming data analysis. To account for heterogeneity, a personalized model is constructed for each data source, and a novel ``subgroup" assumption is employed to capture potential similarities, thereby enhancing model performance. We adopt the penalized renewable estimation method and the efficient proximal gradient descent for model training. The proposed method aligns with both federated and online learning frameworks: raw data are not exchanged among sources, ensuring data privacy, and only summary statistics of previous data batches are required for model updates, significantly reducing storage demands. Theoretically, we establish the consistency properties for model estimation, variable selection, and subgroup structure recovery, demonstrating optimal statistical efficiency. Simulations illustrate the effectiveness of the proposed method. Furthermore, when applied to the financial lending data and the web log data, the proposed method also exhibits advantageous prediction performance. Results of the analysis also provide some practical insights.
Authors: Kyu-Hwan Lee, Seewoo Lee
Abstract: By applying interpretable machine learning methods such as decision trees, we study how simple models can classify the Galois groups of Galois extensions over $\mathbb{Q}$ of degrees 4, 6, 8, 9, and 10, using Dedekind zeta coefficients. Our interpretation of the machine learning results allows us to understand how the distribution of zeta coefficients depends on the Galois group, and to prove new criteria for classifying the Galois groups of these extensions. Combined with previous results, this work provides another example of a new paradigm in mathematical research driven by machine learning.
Authors: Agada Joseph Oche, Arpan Biswas
Abstract: Large language models (LLMs) have emerged as powerful tools for knowledge-intensive tasks across domains. In materials science, to find novel materials for various energy efficient devices for various real-world applications, requires several time and cost expensive simulations and experiments. In order to tune down the uncharted material search space, minimizing the experimental cost, LLMs can play a bigger role to first provide an accelerated search of promising known material candidates. Furthermore, the integration of LLMs with domain-specific information via retrieval-augmented generation (RAG) is poised to revolutionize how researchers predict materials structures, analyze defects, discover novel compounds, and extract knowledge from literature and databases. In motivation to the potentials of LLMs and RAG in accelerating material discovery, this paper presents a broad and systematic review to examine the recent advancements in applying LLMs and RAG to key materials science problems. We survey state-of-the-art developments in crystal structure prediction, defect analysis, materials discovery, literature mining, database integration, and multi-modal retrieval, highlighting how combining LLMs with external knowledge sources enables new capabilities. We discuss the performance, limitations, and implications of these approaches, and outline future directions for leveraging LLMs to accelerate materials research and discovery for advancement in technologies in the area of electronics, optics, biomedical, and energy storage.
Authors: Matthew Fahrbach, Mehrdad Ghadiri
Abstract: We prove that the classic approximation guarantee for the higher-order singular value decomposition (HOSVD) is tight by constructing a tensor for which HOSVD achieves an approximation ratio of $N/(1+\varepsilon)$, for any $\varepsilon > 0$. This matches the upper bound of De Lathauwer et al. (2000a) and shows that the approximation ratio of HOSVD cannot be improved. Using a more advanced construction, we also prove that the approximation guarantees for the ST-HOSVD algorithm of Vannieuwenhoven et al. (2012) and higher-order orthogonal iteration (HOOI) of De Lathauwer et al. (2000b) are tight by showing that they can achieve their worst-case approximation ratio of $N / (1 + \varepsilon)$, for any $\varepsilon > 0$.
Authors: Md Rezwanul Haque, Md. Milon Islam, S M Taslim Uddin Raju, Hamdi Altaheri, Lobna Nassar, Fakhri Karray
Abstract: Depression is a serious mental health illness that significantly affects an individual's well-being and quality of life, making early detection crucial for adequate care and treatment. Detecting depression is often difficult, as it is based primarily on subjective evaluations during clinical interviews. Hence, the early diagnosis of depression, thanks to the content of social networks, has become a prominent research area. The extensive and diverse nature of user-generated information poses a significant challenge, limiting the accurate extraction of relevant temporal information and the effective fusion of data across multiple modalities. This paper introduces MMFformer, a multimodal depression detection network designed to retrieve depressive spatio-temporal high-level patterns from multimodal social media information. The transformer network with residual connections captures spatial features from videos, and a transformer encoder is exploited to design important temporal dynamics in audio. Moreover, the fusion architecture fused the extracted features through late and intermediate fusion strategies to find out the most relevant intermodal correlations among them. Finally, the proposed network is assessed on two large-scale depression detection datasets, and the results clearly reveal that it surpasses existing state-of-the-art approaches, improving the F1-Score by 13.92% for D-Vlog dataset and 7.74% for LMVD dataset. The code is made available publicly at https://github.com/rezwanh001/Large-Scale-Multimodal-Depression-Detection.
URLs: https://github.com/rezwanh001/Large-Scale-Multimodal-Depression-Detection.
Authors: Blair Johnson, Clayton Kerce, Faramarz Fekri
Abstract: Differentiable inductive logic programming (ILP) techniques have proven effective at finding approximate rule-based solutions to link prediction and node classification problems on knowledge graphs; however, the common assumption of chain-like rule structure can hamper the performance and interpretability of existing approaches. We introduce GLIDR, a differentiable rule learning method that models the inference of logic rules with more expressive syntax than previous methods. GLIDR uses a differentiable message passing inference algorithm that generalizes previous chain-like rule learning methods to allow rules with features like branches and cycles. GLIDR has a simple and expressive rule search space which is parameterized by a limit on the maximum number of free variables that may be included in a rule. Explicit logic rules can be extracted from the weights of a GLIDR model for use with symbolic solvers. We demonstrate that GLIDR can significantly outperform existing rule learning methods on knowledge graph completion tasks and even compete with embedding methods despite the inherent disadvantage of being a structure-only prediction method. We show that rules extracted from GLIDR retain significant predictive performance, and that GLIDR is highly robust to training data noise. Finally, we demonstrate that GLIDR can be chained with deep neural networks and optimized end-to-end for rule learning on arbitrary data modalities.
Authors: Yuya Kawakami, Daniel Cayan, Dongyu Liu, Kwan-Liu Ma
Abstract: Ensemble datasets are ever more prevalent in various scientific domains. In climate science, ensemble datasets are used to capture variability in projections under plausible future conditions including greenhouse and aerosol emissions. Each ensemble model run produces projections that are fundamentally similar yet meaningfully distinct. Understanding this variability among ensemble model runs and analyzing its magnitude and patterns is a vital task for climate scientists. In this paper, we present ClimateSOM, a visual analysis workflow that leverages a self-organizing map (SOM) and Large Language Models (LLMs) to support interactive exploration and interpretation of climate ensemble datasets. The workflow abstracts climate ensemble model runs - spatiotemporal time series - into a distribution over a 2D space that captures the variability among the ensemble model runs using a SOM. LLMs are integrated to assist in sensemaking of this SOM-defined 2D space, the basis for the visual analysis tasks. In all, ClimateSOM enables users to explore the variability among ensemble model runs, identify patterns, compare and cluster the ensemble model runs. To demonstrate the utility of ClimateSOM, we apply the workflow to an ensemble dataset of precipitation projections over California and the Northwestern United States. Furthermore, we conduct a short evaluation of our LLM integration, and conduct an expert review of the visual workflow and the insights from the case studies with six domain experts to evaluate our approach and its utility.
Authors: Ngoc N. Tran, Anwar Said, Waseem Abbas, Tyler Derr, Xenofon D. Koutsoukos
Abstract: Graph-based malware classifiers can achieve over 94% accuracy on standard Android datasets, yet we find they suffer accuracy drops of up to 45% when evaluated on previously unseen malware variants from the same family - a scenario where strong generalization would typically be expected. This highlights a key limitation in existing approaches: both the model architectures and their structure-only representations often fail to capture deeper semantic patterns. In this work, we propose a robust semantic enrichment framework that enhances function call graphs with contextual features, including function-level metadata and, when available, code embeddings derived from large language models. The framework is designed to operate under real-world constraints where feature availability is inconsistent, and supports flexible integration of semantic signals. To evaluate generalization under realistic domain and temporal shifts, we introduce two new benchmarks: MalNet-Tiny-Common and MalNet-Tiny-Distinct, constructed using malware family partitioning to simulate cross-family generalization and evolving threat behavior. Experiments across multiple graph neural network backbones show that our method improves classification performance by up to 8% under distribution shift and consistently enhances robustness when integrated with adaptation-based methods. These results offer a practical path toward building resilient malware detection systems in evolving threat environments.
Authors: Alican Yilmaz, Junyang Cai, Serdar Kadioglu, Bistra Dilkina
Abstract: Solving Mixed-Integer Programming (MIP) problems often requires substantial computational resources due to their combinatorial nature. Parallelization has emerged as a critical strategy to accelerate solution times and enhance scalability to tackle large, complex instances. This paper investigates the parallelization capabilities of Balans, a recently proposed multi-armed bandits-based adaptive large neighborhood search for MIPs. While Balans's modular architecture inherently supports parallel exploration of diverse parameter configurations, this potential has not been thoroughly examined. To address this gap, we introduce ParBalans, an extension that leverages both solver-level and algorithmic-level parallelism to improve performance on challenging MIP instances. Our experimental results demonstrate that ParBalans exhibits competitive performance compared to the state-of-the-art commercial solver Gurobi, particularly on hard optimization benchmarks.
Authors: Alejandro Murillo-Gonzalez, Junhong Xu, Lantao Liu
Abstract: Structural causal models describe how the components of a robotic system interact. They provide both structural and functional information about the relationships that are present in the system. The structural information outlines the variables among which there is interaction. The functional information describes how such interactions work, via equations or learned models. In this paper we find that learning the functional relationships while accounting for the uncertainty about the structural information leads to more robust dynamics models which improves downstream planning, while using significantly lower computational resources. This in contrast with common model-learning methods that ignore the causal structure and fail to leverage the sparsity of interactions in robotic systems. We achieve this by estimating a causal structure distribution that is used to sample causal graphs that inform the latent-space representations in an encoder-multidecoder probabilistic model. We show that our model can be used to learn the dynamics of a robot, which together with a sampling-based planner can be used to perform new tasks in novel environments, provided an objective function for the new requirement is available. We validate our method using manipulators and mobile robots in both simulation and the real-world. Additionally, we validate the learned dynamics' adaptability and increased robustness to corrupted inputs and changes in the environment, which is highly desirable in challenging real-world robotics scenarios. Video: https://youtu.be/X6k5t7OOnNc.
Authors: Evangelos Georganas, Dhiraj Kalamkar, Alexander Heinecke
Abstract: The advent of ultra-low-bit LLM models (1/1.58/2-bit), which match the perplexity and end-task performance of their full-precision counterparts using the same model size, is ushering in a new era of LLM inference for resource-constrained environments such as edge devices and AI PCs. While these quantization advances promise models that are more cost-effective in terms of latency, memory, throughput, and energy consumption, the computational efficiency of state-of-the-art (SOTA) inference runtimes (e.g., bitnet.cpp) used to deploy them remains underexplored. In this work, we take a bottom-up approach: we first design and implement 1-bit and 2-bit microkernels optimized for modern CPUs, achieving peak computational efficiency across a variety of CPU platforms. We integrate these microkernels into a state-of-the-art LLM inference framework, namely PyTorch-TPP, and present end-to-end inference results with 2-bit models that outperform the current SOTA runtime bitnet.cpp by up to 2.2x, and deliver up to 7x speedup compared to the 16-bit model inference. Our optimized runtime advances the state of LLM inference on AI PCs and edge devices, paving the way for efficient deployment of ultra-low-bit LLM models.
Authors: Catherine Yeh, Tara Menon, Robin Singh Arya, Helen He, Moira Weigel, Fernanda Vi\'egas, Martin Wattenberg
Abstract: Analyzing literature involves tracking interactions between characters, locations, and themes. Visualization has the potential to facilitate the mapping and analysis of these complex relationships, but capturing structured information from unstructured story data remains a challenge. As large language models (LLMs) continue to advance, we see an opportunity to use their text processing and analysis capabilities to augment and reimagine existing storyline visualization techniques. Toward this goal, we introduce an LLM-driven data parsing pipeline that automatically extracts relevant narrative information from novels and scripts. We then apply this pipeline to create Story Ribbons, an interactive visualization system that helps novice and expert literary analysts explore detailed character and theme trajectories at multiple narrative levels. Through pipeline evaluations and user studies with Story Ribbons on 36 literary works, we demonstrate the potential of LLMs to streamline narrative visualization creation and reveal new insights about familiar stories. We also describe current limitations of AI-based systems, and interaction motifs designed to address these issues.
Authors: Christos Tsirigotis, Vaibhav Adlakha, Joao Monteiro, Aaron Courville, Perouz Taslakian
Abstract: Neural sentence embedding models for dense retrieval typically rely on binary relevance labels, treating query-document pairs as either relevant or irrelevant. However, real-world relevance often exists on a continuum, and recent advances in large language models (LLMs) have made it feasible to scale the generation of fine-grained graded relevance labels. In this work, we propose BiXSE, a simple and effective pointwise training method that optimizes binary cross-entropy (BCE) over LLM-generated graded relevance scores. BiXSE interprets these scores as probabilistic targets, enabling granular supervision from a single labeled query-document pair per query. Unlike pairwise or listwise losses that require multiple annotated comparisons per query, BiXSE achieves strong performance with reduced annotation and compute costs by leveraging in-batch negatives. Extensive experiments across sentence embedding (MMTEB) and retrieval benchmarks (BEIR, TREC-DL) show that BiXSE consistently outperforms softmax-based contrastive learning (InfoNCE), and matches or exceeds strong pairwise ranking baselines when trained on LLM-supervised data. BiXSE offers a robust, scalable alternative for training dense retrieval models as graded relevance supervision becomes increasingly accessible.
Authors: Bowen Zhang, Genan Dai, Hu Huang, Long Lan
Abstract: Graph Neural Networks (GNNs) have demonstrated impressive capabilities in modeling graph-structured data, while Spiking Neural Networks (SNNs) offer high energy efficiency through sparse, event-driven computation. However, existing spiking GNNs predominantly operate in Euclidean space and rely on fixed geometric assumptions, limiting their capacity to model complex graph structures such as hierarchies and cycles. To overcome these limitations, we propose \method{}, a novel Geometry-Aware Spiking Graph Neural Network that unifies spike-based neural dynamics with adaptive representation learning on Riemannian manifolds. \method{} features three key components: a Riemannian Embedding Layer that projects node features into a pool of constant-curvature manifolds, capturing non-Euclidean structures; a Manifold Spiking Layer that models membrane potential evolution and spiking behavior in curved spaces via geometry-consistent neighbor aggregation and curvature-based attention; and a Manifold Learning Objective that enables instance-wise geometry adaptation through jointly optimized classification and link prediction losses defined over geodesic distances. All modules are trained using Riemannian SGD, eliminating the need for backpropagation through time. Extensive experiments on multiple benchmarks show that GSG achieves superior accuracy, robustness, and energy efficiency compared to both Euclidean SNNs and manifold-based GNNs, establishing a new paradigm for curvature-aware, energy-efficient graph learning.
Authors: Benjamin Laufer, Hamidah Oderinwale, Jon Kleinberg
Abstract: Many have observed that the development and deployment of generative machine learning (ML) and artificial intelligence (AI) models follow a distinctive pattern in which pre-trained models are adapted and fine-tuned for specific downstream tasks. However, there is limited empirical work that examines the structure of these interactions. This paper analyzes 1.86 million models on Hugging Face, a leading peer production platform for model development. Our study of model family trees -- networks that connect fine-tuned models to their base or parent -- reveals sprawling fine-tuning lineages that vary widely in size and structure. Using an evolutionary biology lens to study ML models, we use model metadata and model cards to measure the genetic similarity and mutation of traits over model families. We find that models tend to exhibit a family resemblance, meaning their genetic markers and traits exhibit more overlap when they belong to the same model family. However, these similarities depart in certain ways from standard models of asexual reproduction, because mutations are fast and directed, such that two `sibling' models tend to exhibit more similarity than parent/child pairs. Further analysis of the directional drifts of these mutations reveals qualitative insights about the open machine learning ecosystem: Licenses counter-intuitively drift from restrictive, commercial licenses towards permissive or copyleft licenses, often in violation of upstream license's terms; models evolve from multi-lingual compatibility towards english-only compatibility; and model cards reduce in length and standardize by turning, more often, to templates and automatically generated text. Overall, this work takes a step toward an empirically grounded understanding of model fine-tuning and suggests that ecological models and methods can yield novel scientific insights.
Authors: Toan Huynh, Ruth Lopez Fajardo, Guannan Zhang, Lili Ju, Feng Bao
Abstract: We propose a novel framework for adaptively learning the time-evolving solutions of stochastic partial differential equations (SPDEs) using score-based diffusion models within a recursive Bayesian inference setting. SPDEs play a central role in modeling complex physical systems under uncertainty, but their numerical solutions often suffer from model errors and reduced accuracy due to incomplete physical knowledge and environmental variability. To address these challenges, we encode the governing physics into the score function of a diffusion model using simulation data and incorporate observational information via a likelihood-based correction in a reverse-time stochastic differential equation. This enables adaptive learning through iterative refinement of the solution as new data becomes available. To improve computational efficiency in high-dimensional settings, we introduce the ensemble score filter, a training-free approximation of the score function designed for real-time inference. Numerical experiments on benchmark SPDEs demonstrate the accuracy and robustness of the proposed method under sparse and noisy observations.
Authors: Lam Ngo, Huong Ha, Jeffrey Chan, Hongyu Zhang
Abstract: High-dimensional Bayesian Optimization (BO) has attracted significant attention in recent research. However, existing methods have mainly focused on optimizing in continuous domains, while combinatorial (ordinal and categorical) and mixed domains still remain challenging. In this paper, we first propose MOCA-HESP, a novel high-dimensional BO method for combinatorial and mixed variables. The key idea is to leverage the hyper-ellipsoid space partitioning (HESP) technique with different categorical encoders to work with high-dimensional, combinatorial and mixed spaces, while adaptively selecting the optimal encoders for HESP using a multi-armed bandit technique. Our method, MOCA-HESP, is designed as a \textit{meta-algorithm} such that it can incorporate other combinatorial and mixed BO optimizers to further enhance the optimizers' performance. Finally, we develop three practical BO methods by integrating MOCA-HESP with state-of-the-art BO optimizers for combinatorial and mixed variables: standard BO, CASMOPOLITAN, and Bounce. Our experimental results on various synthetic and real-world benchmarks show that our methods outperform existing baselines. Our code implementation can be found at https://github.com/LamNgo1/moca-hesp
Authors: Hamidreza Asadian-Rad, Hossein Soleimani, Shahrokh Farahmand
Abstract: Unmanned aerial vehicles (UAVs) have been recently utilized in multi-access edge computing (MEC) as edge servers. It is desirable to design UAVs' trajectories and user to UAV assignments to ensure satisfactory service to the users and energy efficient operation simultaneously. The posed optimization problem is challenging to solve because: (i) The formulated problem is non-convex, (ii) Due to the mobility of ground users, their future positions and channel gains are not known in advance, (iii) Local UAVs' observations should be communicated to a central entity that solves the optimization problem. The (semi-) centralized processing leads to communication overhead, communication/processing bottlenecks, lack of flexibility and scalability, and loss of robustness to system failures. To simultaneously address all these limitations, we advocate a fully decentralized setup with no centralized entity. Each UAV obtains its local observation and then communicates with its immediate neighbors only. After sharing information with neighbors, each UAV determines its next position via a locally run deep reinforcement learning (DRL) algorithm. None of the UAVs need to know the global communication graph. Two main components of our proposed solution are (i) Graph attention layers (GAT), and (ii) Experience and parameter sharing proximal policy optimization (EPS-PPO). Our proposed approach eliminates all the limitations of semi-centralized MADRL methods such as MAPPO and MA deep deterministic policy gradient (MADDPG), while guaranteeing a better performance than independent local DRLs such as in IPPO. Numerical results reveal notable performance gains in several different criteria compared to the existing MADDPG algorithm, demonstrating the potential for offering a better performance, while utilizing local communications only.
Authors: Gangular Singh Irengbam, Nirvash Singh Wahengbam, Lanthoiba Meitei Khumanthem, Paikhomba Oinam
Abstract: This paper presents the development of a Text-to-Speech (TTS) system for the Manipuri language using the Meitei Mayek script. Leveraging Tacotron 2 and HiFi-GAN, we introduce a neural TTS architecture adapted to support tonal phonology and under-resourced linguistic environments. We develop a phoneme mapping for Meitei Mayek to ARPAbet, curate a single-speaker dataset, and demonstrate intelligible and natural speech synthesis, validated through subjective and objective metrics. This system lays the groundwork for linguistic preservation and technological inclusion of Manipuri.
Authors: Alexander Tyurin
Abstract: We study first-order methods for convex optimization problems with functions $f$ satisfying the recently proposed $\ell$-smoothness condition $||\nabla^{2}f(x)|| \le \ell\left(||\nabla f(x)||\right),$ which generalizes the $L$-smoothness and $(L_{0},L_{1})$-smoothness. While accelerated gradient descent AGD is known to reach the optimal complexity $O(\sqrt{L} R / \sqrt{\varepsilon})$ under $L$-smoothness, where $\varepsilon$ is an error tolerance and $R$ is the distance between a starting and an optimal point, existing extensions to $\ell$-smoothness either incur extra dependence on the initial gradient, suffer exponential factors in $L_{1} R$, or require costly auxiliary sub-routines, leaving open whether an AGD-type $O(\sqrt{\ell(0)} R / \sqrt{\varepsilon})$ rate is possible for small-$\varepsilon$, even in the $(L_{0},L_{1})$-smoothness case. We resolve this open question. Leveraging a new Lyapunov function and designing new algorithms, we achieve $O(\sqrt{\ell(0)} R / \sqrt{\varepsilon})$ oracle complexity for small-$\varepsilon$ and virtually any $\ell$. For instance, for $(L_{0},L_{1})$-smoothness, our bound $O(\sqrt{L_0} R / \sqrt{\varepsilon})$ is provably optimal in the small-$\varepsilon$ regime and removes all non-constant multiplicative factors present in prior accelerated algorithms.
Authors: Giovanni Varricchione, Toryn Q. Klassen, Natasha Alechina, Mehdi Dastani, Brian Logan, Sheila A. McIlraith
Abstract: Reward machines (RMs) are automata structures that encode (non-Markovian) reward functions for reinforcement learning (RL). RMs can reward any behaviour representable in regular languages and, when paired with RL algorithms that exploit RM structure, have been shown to significantly improve sample efficiency in many domains. In this work, we present pushdown reward machines (pdRMs), an extension of reward machines based on deterministic pushdown automata. pdRMs can recognize and reward temporally extended behaviours representable in deterministic context-free languages, making them more expressive than reward machines. We introduce two variants of pdRM-based policies, one which has access to the entire stack of the pdRM, and one which can only access the top $k$ symbols (for a given constant $k$) of the stack. We propose a procedure to check when the two kinds of policies (for a given environment, pdRM, and constant $k$) achieve the same optimal expected reward. We then provide theoretical results establishing the expressive power of pdRMs, and space complexity results about the proposed learning problems. Finally, we provide experimental results showing how agents can be trained to perform tasks representable in deterministic context-free languages using pdRMs.
Authors: Morteza Kimiaei, Vyacheslav Kungurtsev, Brian Olimba
Abstract: Integer and mixed-integer nonlinear programming (INLP, MINLP) are central to logistics, energy, and scheduling, but remain computationally challenging. This survey examines how machine learning and reinforcement learning can enhance exact optimization methods - particularly branch-and-bound (BB), without compromising global optimality. We cover discrete, continuous, and mixed-integer formulations, and highlight applications such as crew scheduling, vehicle routing, and hydropower planning. We introduce a unified BB framework that embeds learning-based strategies into branching, cut selection, node ordering, and parameter control. Classical algorithms are augmented using supervised, imitation, and reinforcement learning models to accelerate convergence while maintaining correctness. We conclude with a taxonomy of learning methods by solver class and learning paradigm, and outline open challenges in generalization, hybridization, and scaling intelligent solvers.
Authors: Siyuan Li, Xi Lin, Guangyan Li, Zehao Liu, Aodu Wulianghai, Li Ding, Jun Wu, Jianhua Li
Abstract: The rapid advancement of large language models (LLMs) has resulted in increasingly sophisticated AI-generated content, posing significant challenges in distinguishing LLM-generated text from human-written language. Existing detection methods, primarily based on lexical heuristics or fine-tuned classifiers, often suffer from limited generalizability and are vulnerable to paraphrasing, adversarial perturbations, and cross-domain shifts. In this work, we propose SentiDetect, a model-agnostic framework for detecting LLM-generated text by analyzing the divergence in sentiment distribution stability. Our method is motivated by the empirical observation that LLM outputs tend to exhibit emotionally consistent patterns, whereas human-written texts display greater emotional variability. To capture this phenomenon, we define two complementary metrics: sentiment distribution consistency and sentiment distribution preservation, which quantify stability under sentiment-altering and semantic-preserving transformations. We evaluate SentiDetect on five diverse datasets and a range of advanced LLMs,including Gemini-1.5-Pro, Claude-3, GPT-4-0613, and LLaMa-3.3. Experimental results demonstrate its superiority over state-of-the-art baselines, with over 16% and 11% F1 score improvements on Gemini-1.5-Pro and GPT-4-0613, respectively. Moreover, SentiDetect also shows greater robustness to paraphrasing, adversarial attacks, and text length variations, outperforming existing detectors in challenging scenarios.
Authors: Wangyue Lu, Lun Du, Sirui Li, Ke Weng, Haozhe Sun, Hengyu Liu, Minghe Yu, Tiancheng Zhang, Ge Yu
Abstract: Interactive theorem provers (ITPs) require manual formalization, which is labor-intensive and demands expert knowledge. While automated formalization offers a potential solution, it faces two major challenges: model hallucination (e.g., undefined predicates, symbol misuse, and version incompatibility) and the semantic gap caused by ambiguous or missing premises in natural language descriptions. To address these issues, we propose CRAMF, a Concept-driven Retrieval-Augmented Mathematical Formalization framework. CRAMF enhances LLM-based autoformalization by retrieving formal definitions of core mathematical concepts, providing contextual grounding during code generation. However, applying retrieval-augmented generation (RAG) in this setting is non-trivial due to the lack of structured knowledge bases, the polymorphic nature of mathematical concepts, and the high precision required in formal retrieval. We introduce a framework for automatically constructing a concept-definition knowledge base from Mathlib4, the standard mathematical library for the Lean 4 theorem prover, indexing over 26,000 formal definitions and 1,000+ core mathematical concepts. To address conceptual polymorphism, we propose contextual query augmentation with domain- and application-level signals. In addition, we design a dual-channel hybrid retrieval strategy with reranking to ensure accurate and relevant definition retrieval. Experiments on miniF2F, ProofNet, and our newly proposed AdvancedMath benchmark show that CRAMF can be seamlessly integrated into LLM-based autoformalizers, yielding consistent improvements in translation accuracy, achieving up to 62.1% and an average of 29.9% relative improvement.
Authors: Hiba Najjar, Deepak Pathak, Marlon Nuske, Andreas Dengel
Abstract: Multimodal learning enables various machine learning tasks to benefit from diverse data sources, effectively mimicking the interplay of different factors in real-world applications, particularly in agriculture. While the heterogeneous nature of involved data modalities may necessitate the design of complex architectures, the model interpretability is often overlooked. In this study, we leverage the intrinsic explainability of Transformer-based models to explain multimodal learning networks, focusing on the task of crop yield prediction at the subfield level. The large datasets used cover various crops, regions, and years, and include four different input modalities: multispectral satellite and weather time series, terrain elevation maps and soil properties. Based on the self-attention mechanism, we estimate feature attributions using two methods, namely the Attention Rollout (AR) and Generic Attention (GA), and evaluate their performance against Shapley-based model-agnostic estimations, Shapley Value Sampling (SVS). Additionally, we propose the Weighted Modality Activation (WMA) method to assess modality attributions and compare it with SVS attributions. Our findings indicate that Transformer-based models outperform other architectures, specifically convolutional and recurrent networks, achieving R2 scores that are higher by 0.10 and 0.04 at the subfield and field levels, respectively. AR is shown to provide more robust and reliable temporal attributions, as confirmed through qualitative and quantitative evaluation, compared to GA and SVS values. Information about crop phenology stages was leveraged to interpret the explanation results in the light of established agronomic knowledge. Furthermore, modality attributions revealed varying patterns across the two methods compared.[...]
Authors: Keqiang Guo, Yuheng Zhong, Xin Tong, Jiangbin Lyu, Rui Zhang
Abstract: Accurately predicting beam-level reference signal received power (RSRP) is essential for beam management in dense multi-user wireless networks, yet challenging due to high measurement overhead and fast channel variations. This paper proposes Neural Beam Field (NBF), a hybrid neural-physical framework for efficient and interpretable spatial beam RSRP prediction. Central to our approach is the introduction of the Multi-path Conditional Power Profile (MCPP), which bridges site-specific multipath propagation with antenna/beam configurations via closed-form analytical modeling. We adopt a decoupled ``blackbox-whitebox" design: a Transformer-based deep neural network (DNN) learns the MCPP from sparse user measurements and positions, while a physics-inspired module analytically infers beam RSRP statistics. To improve convergence and adaptivity, we further introduce a Pretrain-and-Calibrate (PaC) strategy that leverages ray-tracing priors and on-site calibration using RSRP data. Extensive simulations results demonstrate that NBF significantly outperforms conventional table-based channel knowledge maps (CKMs) and pure blackbox DNNs in prediction accuracy, training efficiency, and generalization, while maintaining a compact model size. The proposed framework offers a scalable and physically grounded solution for intelligent beam management in next-generation dense wireless networks.
Authors: Changqing Li, Tianlin Li, Xiaohan Zhang, Aishan Liu, Li Pan
Abstract: Large Language Models (LLMs) face persistent and evolving trustworthiness issues, motivating developers to seek automated and flexible repair methods that enable convenient deployment across diverse scenarios. Existing repair methods like supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) are costly and slow, while prompt engineering lacks robustness and scalability. Representation engineering, which steers model behavior by injecting targeted concept vectors during inference, offers a lightweight, training-free alternative. However, current approaches depend on manually crafted samples and fixed steering strategies, limiting automation and adaptability. To overcome these challenges, we propose MASteer, the first end-to-end framework for trustworthiness repair in LLMs based on representation engineering. MASteer integrates two core components: AutoTester, a multi-agent system that generates diverse, high-quality steer samples tailored to developer needs; and AutoRepairer, which constructs adaptive steering strategies with anchor vectors for automated, context-aware strategy selection during inference. Experiments on standard and customized trustworthiness tasks show MASteer consistently outperforms baselines, improving metrics by 15.36% on LLaMA-3.1-8B-Chat and 4.21% on Qwen-3-8B-Chat, while maintaining general model capabilities. MASteer demonstrates strong robustness, generalization, and practical value for scalable, efficient trustworthiness repair.
Authors: Dan Ivanov, Tristan Freiberg, Haruna Isah
Abstract: DSperse is a modular framework for distributed machine learning inference with strategic cryptographic verification. Operating within the emerging paradigm of distributed zero-knowledge machine learning, DSperse avoids the high cost and rigidity of full-model circuitization by enabling targeted verification of strategically chosen subcomputations. These verifiable segments, or "slices", may cover part or all of the inference pipeline, with global consistency enforced through audit, replication, or economic incentives. This architecture supports a pragmatic form of trust minimization, localizing zero-knowledge proofs to the components where they provide the greatest value. We evaluate DSperse using multiple proving systems and report empirical results on memory usage, runtime, and circuit behavior under sliced and unsliced configurations. By allowing proof boundaries to align flexibly with the model's logical structure, DSperse supports scalable, targeted verification strategies suited to diverse deployment needs.
Authors: M. Adeel Ajaib, Fariha Nasir, Abdul Rehman
Abstract: We explore machine learning techniques for predicting Curie temperatures of magnetic materials using the NEMAD database. By augmenting the dataset with composition-based and domain-aware descriptors, we evaluate the performance of several machine learning models. We find that the Extra Trees Regressor delivers the best performance reaching an R^2 score of up to 0.85 $\pm$ 0.01 (cross-validated) for a balanced dataset. We employ the k-means clustering algorithm to gain insights into the performance of chemically distinct material groups. Furthermore, we perform the SHAP analysis to identify key physicochemical drivers of Curie behavior, such as average atomic number and magnetic moment. By employing explainable AI techniques, this analysis offers insights into the model's predictive behavior, thereby advancing scientific interpretability.
Authors: Myeung Suk Oh, Zhiyao Zhang, FNU Hairi, Alvaro Velasquez, Jia Liu
Abstract: With wireless devices increasingly forming a unified smart network for seamless, user-friendly operations, random access (RA) medium access control (MAC) design is considered a key solution for handling unpredictable data traffic from multiple terminals. However, it remains challenging to design an effective RA-based MAC protocol to minimize collisions and ensure transmission fairness across the devices. While existing multi-agent reinforcement learning (MARL) approaches with centralized training and decentralized execution (CTDE) have been proposed to optimize RA performance, their reliance on centralized training and the significant overhead required for information collection can make real-world applications unrealistic. In this work, we adopt a fully decentralized MARL architecture, where policy learning does not rely on centralized tasks but leverages consensus-based information exchanges across devices. We design our MARL algorithm over an actor-critic (AC) network and propose exchanging only local rewards to minimize communication overhead. Furthermore, we provide a theoretical proof of global convergence for our approach. Numerical experiments show that our proposed MARL algorithm can significantly improve RA network performance compared to other baselines.
Authors: Xintong Chen, Zhenyu Jiang, Jiangbin Lyu, Liqun Fu
Abstract: Intelligent Reflecting Surfaces (IRSs) have potential for significant performance gains in next-generation wireless networks but face key challenges, notably severe double-pathloss and complex multi-user scheduling due to hardware constraints. Active IRSs partially address pathloss but still require efficient scheduling in cell-level multi-IRS multi-user systems, whereby the overhead/delay of channel state acquisition and the scheduling complexity both rise dramatically as the user density and channel dimensions increase. Motivated by these challenges, this paper proposes a novel scheduling framework based on neural Channel Knowledge Map (CKM), designing Transformer-based deep neural networks (DNNs) to predict ergodic spectral efficiency (SE) from historical channel/throughput measurements tagged with user positions. Specifically, two cascaded networks, LPS-Net and SE-Net, are designed to predict link power statistics (LPS) and ergodic SE accurately. We further propose a low-complexity Stable Matching-Iterative Balancing (SM-IB) scheduling algorithm. Numerical evaluations verify that the proposed neural CKM significantly enhances prediction accuracy and computational efficiency, while the SM-IB algorithm effectively achieves near-optimal max-min throughput with greatly reduced complexity.
Authors: Tanjim Bin Faruk, Abdul Matin, Shrideep Pallickara, Sangmi Lee Pallickara
Abstract: Hyperspectral satellite imagery offers sub-30 m views of Earth in hundreds of contiguous spectral bands, enabling fine-grained mapping of soils, crops, and land cover. While self-supervised Masked Autoencoders excel on RGB and low-band multispectral data, they struggle to exploit the intricate spatial-spectral correlations in 200+ band hyperspectral images. We introduce TerraMAE, a novel HSI encoding framework specifically designed to learn highly representative spatial-spectral embeddings for diverse geospatial analyses. TerraMAE features an adaptive channel grouping strategy, based on statistical reflectance properties to capture spectral similarities, and an enhanced reconstruction loss function that incorporates spatial and spectral quality metrics. We demonstrate TerraMAE's effectiveness through superior spatial-spectral information preservation in high-fidelity image reconstruction. Furthermore, we validate its practical utility and the quality of its learned representations through strong performance on three key downstream geospatial tasks: crop identification, land cover classification, and soil texture prediction.
Authors: Shengtao Wen, Haodong Chen, Yadong Wang, Zhongying Pan, Xiang Chen, Yu Tian, Bo Qian, Dong Liang, Sheng-Jun Huang
Abstract: Knowledge editing (KE) provides a scalable approach for updating factual knowledge in large language models without full retraining. While previous studies have demonstrated effectiveness in general domains and medical QA tasks, little attention has been paid to KE in multimodal medical scenarios. Unlike text-only settings, medical KE demands integrating updated knowledge with visual reasoning to support safe and interpretable clinical decisions. To address this gap, we propose MultiMedEdit, the first benchmark tailored to evaluating KE in clinical multimodal tasks. Our framework spans both understanding and reasoning task types, defines a three-dimensional metric suite (reliability, generality, and locality), and supports cross-paradigm comparisons across general and domain-specific models. We conduct extensive experiments under single-editing and lifelong-editing settings. Results suggest that current methods struggle with generalization and long-tail reasoning, particularly in complex clinical workflows. We further present an efficiency analysis (e.g., edit latency, memory footprint), revealing practical trade-offs in real-world deployment across KE paradigms. Overall, MultiMedEdit not only reveals the limitations of current approaches but also provides a solid foundation for developing clinically robust knowledge editing techniques in the future.
Authors: Taeyoun Kwon, Junhyuk Ahn, Taegeun Yun, Heeju Jwa, Yoonchae Choi, Siwon Park, Nam-Joon Kim, Jangchan Kim, Hyun Gon Ryu, Hyuk-Jae Lee
Abstract: Fast Automatic Speech Recognition (ASR) is critical for latency-sensitive applications such as real-time captioning and meeting transcription. However, truly parallel ASR decoding remains challenging due to the sequential nature of autoregressive (AR) decoders and the context limitations of non-autoregressive (NAR) methods. While modern ASR encoders can process up to 30 seconds of audio at once, AR decoders still generate tokens sequentially, creating a latency bottleneck. We propose Whisfusion, the first framework to fuse a pre-trained Whisper encoder with a text diffusion decoder. This NAR architecture resolves the AR latency bottleneck by processing the entire acoustic context in parallel at every decoding step. A lightweight cross-attention adapter trained via parameter-efficient fine-tuning (PEFT) bridges the two modalities. We also introduce a batch-parallel, multi-step decoding strategy that improves accuracy by increasing the number of candidates with minimal impact on speed. Fine-tuned solely on LibriSpeech (960h), Whisfusion achieves a lower WER than Whisper-tiny (8.3% vs. 9.7%), and offers comparable latency on short audio. For longer utterances (>20s), it is up to 2.6x faster than the AR baseline, establishing a new, efficient operating point for long-form ASR. The implementation and training scripts are available at https://github.com/taeyoun811/Whisfusion.
Authors: Tran Tuan Kiet, Nguyen Thang Loi, Vo Nguyen Le Duy
Abstract: Anomaly detection (AD) plays a vital role across a wide range of domains, but its performance might deteriorate when applied to target domains with limited data. Domain Adaptation (DA) offers a solution by transferring knowledge from a related source domain with abundant data. However, this adaptation process can introduce additional uncertainty, making it difficult to draw statistically valid conclusions from AD results. In this paper, we propose STAND-DA -- a novel framework for statistically rigorous Autoencoder-based AD after Representation Learning-based DA. Built on the Selective Inference (SI) framework, STAND-DA computes valid $p$-values for detected anomalies and rigorously controls the false positive rate below a pre-specified level $\alpha$ (e.g., 0.05). To address the computational challenges of applying SI to deep learning models, we develop the GPU-accelerated SI implementation, significantly enhancing both scalability and runtime performance. This advancement makes SI practically feasible for modern, large-scale deep architectures. Extensive experiments on synthetic and real-world datasets validate the theoretical results and computational efficiency of the proposed STAND-DA method.
Authors: Wenhan Liu, Xinyu Ma, Weiwei Sun, Yutao Zhu, Yuchen Li, Dawei Yin, Zhicheng Dou
Abstract: Large Language Model (LLM) based listwise ranking has shown superior performance in many passage ranking tasks. With the development of Large Reasoning Models, many studies have demonstrated that step-by-step reasoning during test-time helps improve listwise ranking performance. However, due to the scarcity of reasoning-intensive training data, existing rerankers perform poorly in many complex ranking scenarios and the ranking ability of reasoning-intensive rerankers remains largely underdeveloped. In this paper, we first propose an automated reasoning-intensive training data synthesis framework, which sources training queries and passages from diverse domains and applies DeepSeek-R1 to generate high-quality training labels. A self-consistency data filtering mechanism is designed to ensure the data quality. To empower the listwise reranker with strong reasoning ability, we further propose a two-stage post-training approach, which includes a cold-start supervised fine-tuning (SFT) stage for reasoning pattern learning and a reinforcement learning (RL) stage for further ranking ability enhancement. During the RL stage, based on the nature of listwise ranking, we design a multi-view ranking reward, which is more effective than a ranking metric-based reward. Extensive experiments demonstrate that our trained reasoning-intensive reranker \textbf{ReasonRank} outperforms existing baselines significantly and also achieves much lower latency than pointwise reranker Rank1. \textbf{Through further experiments, our ReasonRank has achieved state-of-the-art (SOTA) performance 40.6 on the BRIGHT leaderboard\footnote{https://brightbenchmark.github.io/}.} Our codes are available at https://github.com/8421BCD/ReasonRank.
URLs: https://brightbenchmark.github.io/, https://github.com/8421BCD/ReasonRank.
Authors: Jason C. Furtado, Maria J. Molina, Marybeth C. Arcodia, Weston Anderson, Tom Beucler, John A. Callahan, Laura M. Ciasto, Vittorio A. Gensini, Michelle L'Heureux, Kathleen Pegion, Jhayron S. P\'erez-Carrasquilla, Maike Sonnewald, Ken Takahashi, Baoqiang Xiang, Brian G. Zimmerman
Abstract: Artificial intelligence (AI) -- and specifically machine learning (ML) -- applications for climate prediction across timescales are proliferating quickly. The emergence of these methods prompts a revisit to the impact of data preprocessing, a topic familiar to the climate community, as more traditional statistical models work with relatively small sample sizes. Indeed, the skill and confidence in the forecasts produced by data-driven models are directly influenced by the quality of the datasets and how they are treated during model development, thus yielding the colloquialism "garbage in, garbage out." As such, this article establishes protocols for the proper preprocessing of input data for AI/ML models designed for climate prediction (i.e., subseasonal to decadal and longer). The three aims are to: (1) educate researchers, developers, and end users on the effects that preprocessing has on climate predictions; (2) provide recommended practices for data preprocessing for such applications; and (3) empower end users to decipher whether the models they are using are properly designed for their objectives. Specific topics covered in this article include the creation of (standardized) anomalies, dealing with non-stationarity and the spatiotemporally correlated nature of climate data, and handling of extreme values and variables with potentially complex distributions. Case studies will illustrate how using different preprocessing techniques can produce different predictions from the same model, which can create confusion and decrease confidence in the overall process. Ultimately, implementing the recommended practices set forth in this article will enhance the robustness and transparency of AI/ML in climate prediction studies.
Authors: Haodi Jiang, Qin Li, Jason T. L. Wang, Haimin Wang, Serena Criscuoli
Abstract: Solar extreme ultraviolet (EUV) irradiance plays a crucial role in heating the Earth's ionosphere, thermosphere, and mesosphere, affecting atmospheric dynamics over varying time scales. Although significant effort has been spent studying short-term EUV variations from solar transient events, there is little work to explore the long-term evolution of the EUV flux over multiple solar cycles. Continuous EUV flux measurements have only been available since 1995, leaving significant gaps in earlier data. In this study, we propose a Bayesian deep learning model, named SEMNet, to fill the gaps. We validate our approach by applying SEMNet to construct SOHO/SEM EUV flux measurements in the period between 1998 and 2014 using CaII K images from the Precision Solar Photometric Telescope. We then extend SEMNet through transfer learning to reconstruct solar EUV irradiance in the period between 1950 and 1960 using CaII K images from the Kodaikanal Solar Observatory. Experimental results show that SEMNet provides reliable predictions along with uncertainty bounds, demonstrating the feasibility of CaII K images as a robust proxy for long-term EUV fluxes. These findings contribute to a better understanding of solar influences on Earth's climate over extended periods.
Authors: Chenxu Zhao, Wei Qian, Aobo Chen, Mengdi Huai
Abstract: Recent studies have shown that deep learning models are vulnerable to membership inference attacks (MIAs), which aim to infer whether a data record was used to train a target model or not. To analyze and study these vulnerabilities, various MIA methods have been proposed. Despite the significance and popularity of MIAs, existing works on MIAs are limited in providing guarantees on the false discovery rate (FDR), which refers to the expected proportion of false discoveries among the identified positive discoveries. However, it is very challenging to ensure the false discovery rate guarantees, because the underlying distribution is usually unknown, and the estimated non-member probabilities often exhibit interdependence. To tackle the above challenges, in this paper, we design a novel membership inference attack method, which can provide the guarantees on the false discovery rate. Additionally, we show that our method can also provide the marginal probability guarantee on labeling true non-member data as member data. Notably, our method can work as a wrapper that can be seamlessly integrated with existing MIA methods in a post-hoc manner, while also providing the FDR control. We perform the theoretical analysis for our method. Extensive experiments in various settings (e.g., the black-box setting and the lifelong learning setting) are also conducted to verify the desirable performance of our method.
Authors: Beilong Tang, Xiaoxiao Miao, Xin Wang, Ming Li
Abstract: Voice anonymization protects speaker privacy by concealing identity while preserving linguistic and paralinguistic content. Self-supervised learning (SSL) representations encode linguistic features but preserve speaker traits. We propose a novel speaker-embedding-free framework called SEF-MK. Instead of using a single k-means model trained on the entire dataset, SEF-MK anonymizes SSL representations for each utterance by randomly selecting one of multiple k-means models, each trained on a different subset of speakers. We explore this approach from both attacker and user perspectives. Extensive experiments show that, compared to a single k-means model, SEF-MK with multiple k-means models better preserves linguistic and emotional content from the user's viewpoint. However, from the attacker's perspective, utilizing multiple k-means models boosts the effectiveness of privacy attacks. These insights can aid users in designing voice anonymization systems to mitigate attacker threats.
Authors: Yaswitha Gujju, Romain Harang, Chao Li, Tetsuo Shibuya, Qibin Zhao
Abstract: The quest for effective quantum feature maps for data encoding presents significant challenges, particularly due to the flat training landscapes and lengthy training processes associated with parameterised quantum circuits. To address these issues, we propose an evolutionary training-free quantum architecture search (QAS) framework that employs circuit-based heuristics focused on trainability, hardware robustness, generalisation ability, expressivity, complexity, and kernel-target alignment. By ranking circuit architectures with various proxies, we reduce evaluation costs and incorporate hardware-aware circuits to enhance robustness against noise. We evaluate our approach on classification tasks (using quantum support vector machine) across diverse datasets using both artificial and quantum-generated datasets. Our approach demonstrates competitive accuracy on both simulators and real quantum hardware, surpassing state-of-the-art QAS methods in terms of sampling efficiency and achieving up to a 2x speedup in architecture search runtime.
Authors: Nikolai Warner, Wenjin Zhang, Irfan Essa, Apaar Sadhwani
Abstract: Lifting-based methods for 3D Human Pose Estimation (HPE), which predict 3D poses from detected 2D keypoints, often generalize poorly to new datasets and real-world settings. To address this, we propose \emph{AugLift}, a simple yet effective reformulation of the standard lifting pipeline that significantly improves generalization performance without requiring additional data collection or sensors. AugLift sparsely enriches the standard input -- the 2D keypoint coordinates $(x, y)$ -- by augmenting it with a keypoint detection confidence score $c$ and a corresponding depth estimate $d$. These additional signals are computed from the image using off-the-shelf, pre-trained models (e.g., for monocular depth estimation), thereby inheriting their strong generalization capabilities. Importantly, AugLift serves as a modular add-on and can be readily integrated into existing lifting architectures. Our extensive experiments across four datasets demonstrate that AugLift boosts cross-dataset performance on unseen datasets by an average of $10.1\%$, while also improving in-distribution performance by $4.0\%$. These gains are consistent across various lifting architectures, highlighting the robustness of our method. Our analysis suggests that these sparse, keypoint-aligned cues provide robust frame-level context, offering a practical way to significantly improve the generalization of any lifting-based pose estimation model. Code will be made publicly available.
Authors: Antonino Greco, Marco D'Alessandro, Karl J. Friston, Giovanni Pezzulo, Markus Siegel
Abstract: Biological systems leverage top-down feedback for visual processing, yet most artificial vision models succeed in image classification using purely feedforward or recurrent architectures, calling into question the functional significance of descending cortical pathways. Here, we trained convolutional recurrent neural networks (ConvRNN) on image classification in the presence or absence of top-down feedback projections to elucidate the specific computational contributions of those feedback pathways. We found that ConvRNNs with top-down feedback exhibited remarkable speed-accuracy trade-off and robustness to noise perturbations and adversarial attacks, but only when they were trained with stochastic neural variability, simulated by randomly silencing single units via dropout. By performing detailed analyses to identify the reasons for such benefits, we observed that feedback information substantially shaped the representational geometry of the post-integration layer, combining the bottom-up and top-down streams, and this effect was amplified by dropout. Moreover, feedback signals coupled with dropout optimally constrained network activity onto a low-dimensional manifold and encoded object information more efficiently in out-of-distribution regimes, with top-down information stabilizing the representational dynamics at the population level. Together, these findings uncover a dual mechanism for resilient sensory coding. On the one hand, neural stochasticity prevents unit-level co-adaptation albeit at the cost of more chaotic dynamics. On the other hand, top-down feedback harnesses high-level information to stabilize network activity on compact low-dimensional manifolds.
Authors: Ahmed M. Abuzuraiq, Philippe Pasquier
Abstract: Explainable AI (XAI) in creative contexts can go beyond transparency to support artistic engagement, modifiability, and sustained practice. While curated datasets and training human-scale models can offer artists greater agency and control, large-scale generative models like text-to-image diffusion systems often obscure these possibilities. We suggest that even large models can be treated as creative materials if their internal structure is exposed and manipulable. We propose a craft-based approach to explainability rooted in long-term, hands-on engagement akin to Sch\"on's "reflection-in-action" and demonstrate its application through a model-bending and inspection plugin integrated into the node-based interface of ComfyUI. We demonstrate that by interactively manipulating different parts of a generative model, artists can develop an intuition about how each component influences the output.
Authors: Kabir Khan, Priya Sharma, Arjun Mehta, Neha Gupta, Ravi Narayanan
Abstract: Large Language Models (LLMs) suffer from a critical limitation: their knowledge is static and quickly becomes outdated. Retraining these massive models is computationally prohibitive, while existing knowledge editing techniques can be slow and may introduce unforeseen side effects. To address this, we propose DySK-Attn, a novel framework that enables LLMs to efficiently integrate real-time knowledge from a dynamic external source. Our approach synergizes an LLM with a dynamic Knowledge Graph (KG) that can be updated instantaneously. The core of our framework is a sparse knowledge attention mechanism, which allows the LLM to perform a coarse-to-fine grained search, efficiently identifying and focusing on a small, highly relevant subset of facts from the vast KG. This mechanism avoids the high computational cost of dense attention over the entire knowledge base and mitigates noise from irrelevant information. We demonstrate through extensive experiments on time-sensitive question-answering tasks that DySK-Attn significantly outperforms strong baselines, including standard Retrieval-Augmented Generation (RAG) and model editing techniques, in both factual accuracy for updated knowledge and computational efficiency. Our framework offers a scalable and effective solution for building LLMs that can stay current with the ever-changing world.
Authors: Itai Allouche, Itay Asael, Rotem Rousso, Vered Dassa, Ann Bradlow, Seung-Eun Kim, Matthew Goldrick, Joseph Keshet
Abstract: Despite their success in speech processing, neural networks often operate as black boxes, prompting the question: what informs their decisions, and how can we interpret them? This work examines this issue in the context of lexical stress. A dataset of English disyllabic words was automatically constructed from read and spontaneous speech. Several Convolutional Neural Network (CNN) architectures were trained to predict stress position from a spectrographic representation of disyllabic words lacking minimal stress pairs (e.g., initial stress WAllet, final stress exTEND), achieving up to 92% accuracy on held-out test data. Layerwise Relevance Propagation (LRP), a technique for CNN interpretability analysis, revealed that predictions for held-out minimal pairs (PROtest vs. proTEST ) were most strongly influenced by information in stressed versus unstressed syllables, particularly the spectral properties of stressed vowels. However, the classifiers also attended to information throughout the word. A feature-specific relevance analysis is proposed, and its results suggest that our best-performing classifier is strongly influenced by the stressed vowel's first and second formants, with some evidence that its pitch and third formant also contribute. These results reveal deep learning's ability to acquire distributed cues to stress from naturally occurring data, extending traditional phonetic work based around highly controlled stimuli.
Authors: Raunak Narwal, Syed Abbas
Abstract: Modelling disease outbreak models remains challenging due to incomplete surveillance data, noise, and limited access to standardized datasets. We have created BIGBOY1.2, an open synthetic dataset generator that creates configurable epidemic time series and population-level trajectories suitable for benchmarking modelling, forecasting, and visualisation. The framework supports SEIR and SIR-like compartmental logic, custom seasonality, and noise injection to mimic real reporting artifacts. BIGBOY1.2 can produce datasets with diverse characteristics, making it suitable for comparing traditional epidemiological models (e.g., SIR, SEIR) with modern machine learning approaches (e.g., SVM, neural networks).
Authors: Xiang Xiang, Qinhao Zhou, Zhuo Xu, Jing Ma, Jiaxin Dai, Yifan Liang, Hanlin Li
Abstract: Substantial progress has been made in various techniques for open-world recognition. Out-of-distribution (OOD) detection methods can effectively distinguish between known and unknown classes in the data, while incremental learning enables continuous model knowledge updates. However, in open-world scenarios, these approaches still face limitations. Relying solely on OOD detection does not facilitate knowledge updates in the model, and incremental fine-tuning typically requires supervised conditions, which significantly deviate from open-world settings. To address these challenges, this paper proposes OpenHAIV, a novel framework that integrates OOD detection, new class discovery, and incremental continual fine-tuning into a unified pipeline. This framework allows models to autonomously acquire and update knowledge in open-world environments. The proposed framework is available at https://haiv-lab.github.io/openhaiv .
Authors: Mahdi Maleki, Reza Agahzadeh Ayoubi, Marouan Mizmizi, Umberto Spagnolini
Abstract: This paper introduces the use of static electromagnetic skins (EMSs) to enable robust device localization via channel charting (CC) in realistic urban environments. We develop a rigorous optimization framework that leverages EMS to enhance channel dissimilarity and spatial fingerprinting, formulating EMS phase profile design as a codebook-based problem targeting the upper quantiles of key embedding metric, localization error, trustworthiness, and continuity. Through 3D ray-traced simulations of a representative city scenario, we demonstrate that optimized EMS configurations, in addition to significant improvement of the average positioning error, reduce the 90th-percentile localization error from over 60 m (no EMS) to less than 25 m, while drastically improving trustworthiness and continuity. To the best of our knowledge, this is the first work to exploit Smart Radio Environment (SRE) with static EMS for enhancing CC, achieving substantial gains in localization performance under challenging None-Line-of-Sight (NLoS) conditions.
Authors: Cristian Cosentino, Annamaria Defilippo, Marco Dossena, Christopher Irwin, Sara Joubbi, Pietro Li\`o
Abstract: HealthBranches is a novel benchmark dataset for medical Question-Answering (Q&A), specifically designed to evaluate complex reasoning in Large Language Models (LLMs). This dataset is generated through a semi-automated pipeline that transforms explicit decision pathways from medical source into realistic patient cases with associated questions and answers. Covering 4,063 case studies across 17 healthcare topics, each data point is based on clinically validated reasoning chains. HealthBranches supports both open-ended and multiple-choice question formats and uniquely includes the full reasoning path for each Q&A. Its structured design enables robust evaluation of LLMs' multi-step inference capabilities, including their performance in structured Retrieval-Augmented Generation (RAG) contexts. HealthBranches establishes a foundation for the development of more trustworthy, interpretable, and clinically reliable LLMs in high-stakes domains while also serving as a valuable resource for educational purposes.
Authors: Lilit Grigoryan, Vladimir Bataev, Nikolay Karpov, Andrei Andrusenko, Vitaly Lavrukhin, Boris Ginsburg
Abstract: While beam search improves speech recognition quality over greedy decoding, standard implementations are slow, often sequential, and CPU-bound. To fully leverage modern hardware capabilities, we present a novel open-source FlexCTC toolkit for fully GPU-based beam decoding, designed for Connectionist Temporal Classification (CTC) models. Developed entirely in Python and PyTorch, it offers a fast, user-friendly, and extensible alternative to traditional C++, CUDA, or WFST-based decoders. The toolkit features a high-performance, fully batched GPU implementation with eliminated CPU-GPU synchronization and minimized kernel launch overhead via CUDA Graphs. It also supports advanced contextualization techniques, including GPU-powered N-gram language model fusion and phrase-level boosting. These features enable accurate and efficient decoding, making them suitable for both research and production use.
Authors: Shubhra Ghosh, Abhilekh Borah, Aditya Kumar Guru, Kripabandhu Ghosh
Abstract: The rapid proliferation of Large Language Models (LLMs) has significantly contributed to the development of equitable AI systems capable of factual question-answering (QA). However, no known study tests the LLMs' robustness when presented with obfuscated versions of questions. To systematically evaluate these limitations, we propose a novel technique, ObfusQAte and, leveraging the same, introduce ObfusQA, a comprehensive, first of its kind, framework with multi-tiered obfuscation levels designed to examine LLM capabilities across three distinct dimensions: (i) Named-Entity Indirection, (ii) Distractor Indirection, and (iii) Contextual Overload. By capturing these fine-grained distinctions in language, ObfusQA provides a comprehensive benchmark for evaluating LLM robustness and adaptability. Our study observes that LLMs exhibit a tendency to fail or generate hallucinated responses when confronted with these increasingly nuanced variations. To foster research in this direction, we make ObfusQAte publicly available.
Authors: Polina V. Banushkina, Sergei V. Krivov
Abstract: Rare but critical events in complex systems, such as protein folding, chemical reactions, disease progression, and extreme weather or climate phenomena, are governed by complex, high-dimensional, stochastic dynamics. Identifying an optimal reaction coordinate (RC) that accurately captures the progress of these dynamics is crucial for understanding and simulating such processes. This work introduces a nonparametric RC optimization framework that incorporates trajectory histories, enabling robust analysis even for irregular or incomplete data. The power of the method is demonstrated through increasingly challenging analyses of protein folding dynamics, where it provides accurate committor estimates that pass a stringent validation test and yield high-resolution free energy profiles. Its generality is further illustrated through applications to dynamics in phase space, a conceptual ocean circulation model, and a longitudinal clinical dataset. These results demonstrate that rare event dynamics can be accurately characterized without exhaustive sampling of the configuration space, establishing a general, flexible, and robust framework for analyzing complex dynamical systems and longitudinal datasets.
Authors: Rubing Chen, Jiaxin Wu, Jian Wang, Xulu Zhang, Wenqi Fan, Chenghua Lin, Xiao-Yong Wei, Qing Li
Abstract: Numerous benchmarks have been built to evaluate the domain-specific abilities of large language models (LLMs), highlighting the need for effective and efficient benchmark construction. Existing domain-specific benchmarks primarily focus on the scaling law, relying on massive corpora for supervised fine-tuning or generating extensive question sets for broad coverage. However, the impact of corpus and question-answer (QA) set design on the precision and recall of domain-specific LLMs remains unexplored. In this paper, we address this gap and demonstrate that the scaling law is not always the optimal principle for benchmark construction in specific domains. Instead, we propose Comp-Comp, an iterative benchmarking framework based on a comprehensiveness-compactness principle. Here, comprehensiveness ensures semantic recall of the domain, while compactness enhances precision, guiding both corpus and QA set construction. To validate our framework, we conducted a case study in a well-renowned university, resulting in the creation of XUBench, a large-scale and comprehensive closed-domain benchmark. Although we use the academic domain as the case in this work, our Comp-Comp framework is designed to be extensible beyond academia, providing valuable insights for benchmark construction across various domains.
Authors: He Kong, Die Hu, Jingguo Ge, Liangxiong Li, Hui Li, Tong Li
Abstract: Automating penetration testing is crucial for enhancing cybersecurity, yet current Large Language Models (LLMs) face significant limitations in this domain, including poor error handling, inefficient reasoning, and an inability to perform complex end-to-end tasks autonomously. To address these challenges, we introduce Pentest-R1, a novel framework designed to optimize LLM reasoning capabilities for this task through a two-stage reinforcement learning pipeline. We first construct a dataset of over 500 real-world, multi-step walkthroughs, which Pentest-R1 leverages for offline reinforcement learning (RL) to instill foundational attack logic. Subsequently, the LLM is fine-tuned via online RL in an interactive Capture The Flag (CTF) environment, where it learns directly from environmental feedback to develop robust error self-correction and adaptive strategies. Our extensive experiments on the Cybench and AutoPenBench benchmarks demonstrate the framework's effectiveness. On AutoPenBench, Pentest-R1 achieves a 24.2\% success rate, surpassing most state-of-the-art models and ranking second only to Gemini 2.5 Flash. On Cybench, it attains a 15.0\% success rate in unguided tasks, establishing a new state-of-the-art for open-source LLMs and matching the performance of top proprietary models. Ablation studies confirm that the synergy of both training stages is critical to its success.
Authors: Jun Li
Abstract: This work presents a statistical mechanics characterization of neural networks, motivated by the replica symmetry breaking (RSB) phenomenon in spin glasses. A Hopfield-type spin glass model is constructed from a given feedforward neural network (FNN). Overlaps between simulated replica samples serve as a characteristic descriptor of the FNN. The connection between the spin-glass description and commonly studied properties of the FNN -- such as data fitting, capacity, generalization, and robustness -- has been investigated and empirically demonstrated. Unlike prior analytical studies that focus on model ensembles, this method provides a computable descriptor for individual network instances, which reveals nontrivial structural properties that are not captured by conventional metrics such as loss or accuracy. Preliminary results suggests its potential for practical applications such as model inspection, safety verification, and detection of hidden vulnerabilities.
Authors: Jesse Ponnock
Abstract: Given recent breakthroughs in Generative Artificial Intelligence (GAI) and Large Language Models (LLMs), more and more professional services are being augmented through Artificial Intelligence (AI), which once seemed impossible to automate. This paper presents a modular model for leveraging GAI in developing strategic plans for large scale government organizations and evaluates leading machine learning techniques in their application towards one of the identified modules. Specifically, the performance of BERTopic and Non-negative Matrix Factorization (NMF) are evaluated in their ability to use topic modeling to generate themes representative of Vision Elements within a strategic plan. To accomplish this, BERTopic and NMF models are trained using a large volume of reports from the Government Accountability Office (GAO). The generated topics from each model are then scored for similarity against the Vision Elements of a published strategic plan and the results are compared. Our results show that these techniques are capable of generating themes similar to 100% of the elements being evaluated against. Further, we conclude that BERTopic performs best in this application with more than half of its correlated topics achieving a "medium" or "strong" correlation. A capability of GAI-enabled strategic plan development impacts a multi-billion dollar industry and assists the federal government in overcoming regulatory requirements which are crucial to the public good. Further work will focus on the operationalization of the concept proven in this study as well as viability of the remaining modules in the proposed model for GAI-generated strategic plans.
Authors: Yueyi Wang, Qiyao Wei
Abstract: In this study, we wish to showcase the unique utility of large language models (LLMs) in financial semantic annotation and alpha signal discovery. Leveraging a corpus of company-related tweets, we use an LLM to automatically assign multi-label event categories to high-sentiment-intensity tweets. We align these labeled sentiment signals with forward returns over 1-to-7-day horizons to evaluate their statistical efficacy and market tradability. Our experiments reveal that certain event labels consistently yield negative alpha, with Sharpe ratios as low as -0.38 and information coefficients exceeding 0.05, all statistically significant at the 95\% confidence level. This study establishes the feasibility of transforming unstructured social media text into structured, multi-label event variables. A key contribution of this work is its commitment to transparency and reproducibility; all code and methodologies are made publicly available. Our results provide compelling evidence that social media sentiment is a valuable, albeit noisy, signal in financial forecasting and underscore the potential of open-source frameworks to democratize algorithmic trading research.
Authors: Jean de Dieu Nyandwi, Yueqi Song, Simran Khanuja, Graham Neubig
Abstract: Multimodal Large Language Models excel in high-resource settings, but often misinterpret long-tail cultural entities and underperform in low-resource languages. To address this gap, we propose a data-centric approach that directly grounds MLLMs in cultural knowledge. Leveraging a large scale knowledge graph from Wikidata, we collect images that represent culturally significant entities, and generate synthetic multilingual visual question answering data. The resulting dataset, CulturalGround, comprises 22 million high-quality, culturally-rich VQA pairs spanning 42 countries and 39 languages. We train an open-source MLLM CulturalPangea on CulturalGround, interleaving standard multilingual instruction-tuning data to preserve general abilities. CulturalPangea achieves state-of-the-art performance among open models on various culture-focused multilingual multimodal benchmarks, outperforming prior models by an average of 5.0 without degrading results on mainstream vision-language tasks. Our findings show that our targeted, culturally grounded approach could substantially narrow the cultural gap in MLLMs and offer a practical path towards globally inclusive multimodal systems.
Authors: Xinjia Lu, Chuhan Wang, Qian Zhao, Lixing Zhu, Xuehu Zhu
Abstract: We focus on the critical challenge of handling non-stationary data streams in online continual learning environments, where constrained storage capacity prevents complete retention of historical data, leading to catastrophic forgetting during sequential task training. To more effectively analyze and address the problem of catastrophic forgetting in continual learning, we propose a novel continual learning framework from a statistical perspective. Our approach incorporates random effects across all model parameters and allows the dimension of parameters to diverge to infinity, offering a general formulation for continual learning problems. To efficiently process streaming data, we develop a Multi-step Newton Iteration algorithm that significantly reduces computational costs in certain scenarios by alleviating the burden of matrix inversion. Theoretically, we derive the asymptotic normality of the estimator, enabling subsequent statistical inference. Comprehensive validation through synthetic data experiments and two real datasets analyses demonstrates the effectiveness of our proposed method.
Authors: Fotis I. Giasemis
Abstract: As the particle physics community needs higher and higher precisions in order to test our current model of the subatomic world, larger and larger datasets are necessary. With upgrades scheduled for the detectors of colliding-beam experiments around the world, and specifically at the Large Hadron Collider at CERN, more collisions and more complex interactions are expected. This directly implies an increase in data produced and consequently in the computational resources needed to process them. At CERN, the amount of data produced is gargantuan. This is why the data have to be heavily filtered and selected in real time before being permanently stored. This data can then be used to perform physics analyses, in order to expand our current understanding of the universe and improve the Standard Model of physics. This real-time filtering, known as triggering, involves complex processing happening often at frequencies as high as 40 MHz. This thesis contributes to understanding how machine learning models can be efficiently deployed in such environments, in order to maximize throughput and minimize energy consumption. Inevitably, modern hardware designed for such tasks and contemporary algorithms are needed in order to meet the challenges posed by the stringent, high-frequency data rates. In this work, I present our graph neural network-based pipeline, developed for charged particle track reconstruction at the LHCb experiment at CERN. The pipeline was implemented end-to-end inside LHCb's first-level trigger, entirely on GPUs. Its performance was compared against the classical tracking algorithms currently in production at LHCb. The pipeline was also accelerated on the FPGA architecture, and its performance in terms of power consumption and processing speed was compared against the GPU implementation.
Authors: Stefan Szeider
Abstract: Translating natural language problem descriptions into formal constraint models remains a fundamental challenge in constraint programming, requiring deep expertise in both the problem domain and modeling frameworks. Previous approaches to automating this translation have employed fixed workflows with predetermined modeling steps, failing on a significant number of benchmark problems. We present a new approach using a pure agentic strategy without any fixed pipeline. We developed a general-purpose Python coding agent based on the ReAct (Reason and Act) principle, utilizing a persistent IPython kernel for stateful code execution and iterative development. Rather than embedding constraint programming logic into the agent architecture, domain-specific expertise is injected solely through a carefully crafted project prompt. The agent combines this prompt-encoded knowledge with access to file operations and code execution tools, enabling it to test hypotheses, debug failures, and verify solutions dynamically. Implemented in just a few hundred lines of code, this architecture successfully solves all 101 problems of the CP-Bench constraint programming benchmark set. The results suggest that constraint modeling tasks require the combination of general coding tools and domain expertise encoded in prompts, rather than specialized agent architectures or predefined workflows.
Authors: Alexander Duffy, Samuel J Paech, Ishana Shastri, Elizabeth Karpinski, Baptiste Alloui-Cros, Tyler Marques, Matthew Lyle Olson
Abstract: We present the first evaluation harness that enables any out-of-the-box, local, Large Language Models (LLMs) to play full-press Diplomacy without fine-tuning or specialized training. Previous work required frontier LLMs, or fine-tuning, due to the high complexity and information density of Diplomacy's game state. Combined with the high variance of matches, these factors made Diplomacy prohibitive for study. In this work, we used data-driven iteration to optimize a textual game state representation such that a 24B model can reliably complete matches without any fine tuning. We develop tooling to facilitate hypothesis testing and statistical analysis, and we present case studies on persuasion, aggressive playstyles, and performance across a range of models. We conduct a variety of experiments across many popular LLMs, finding the larger models perform the best, but the smaller models still play adequately. We also introduce Critical State Analysis: an experimental protocol for rapidly iterating and analyzing key moments in a game at depth. Our harness democratizes the evaluation of strategic reasoning in LLMs by eliminating the need for fine-tuning, and it provides insights into how these capabilities emerge naturally from widely used LLMs. Our code is available in the supplement and will be open sourced.
Authors: Vukan Ninkovic, Dejan Vukobratovic
Abstract: Unequal error protection (UEP) coding that enables differentiated reliability levels within a transmitted message is essential for modern communication systems. Autoencoder (AE)-based code designs have shown promise in the context of learned equal error protection (EEP) coding schemes. However, their application to UEP remains largely unexplored, particularly at intermediate blocklengths, due to the increasing complexity of AE-based models. Inspired by the proven effectiveness of superposition coding and successive interference cancellation (SIC) decoding in conventional UEP schemes, we propose a structured AE-based architecture that extends AE-based UEP codes to substantially larger blocklengths while maintaining efficient training. By structuring encoding and decoding into smaller AE subblocks, our method provides a flexible framework for fine-tuning UEP reliability levels while adapting to diverse system parameters. Numerical results show that the proposed approach improves over established achievability bounds of randomized superposition coding-based UEP schemes with SIC decoding, making the proposed structured AE-based UEP codes a scalable and efficient solution for next-generation networks.
Authors: Chidaksh Ravuru
Abstract: Automated soccer commentary generation has evolved from template-based systems to advanced neural architectures, aiming to produce real-time descriptions of sports events. While frameworks like SoccerNet-Caption laid foundational work, their inability to achieve fine-grained alignment between video content and commentary remains a significant challenge. Recent efforts such as MatchTime, with its MatchVoice model, address this issue through coarse and fine-grained alignment techniques, achieving improved temporal synchronization. In this paper, we extend MatchVoice to commentary generation for soccer highlights using the GOAL dataset, which emphasizes short clips over entire games. We conduct extensive experiments to reproduce the original MatchTime results and evaluate our setup, highlighting the impact of different training configurations and hardware limitations. Furthermore, we explore the effect of varying window sizes on zero-shot performance. While MatchVoice exhibits promising generalization capabilities, our findings suggest the need for integrating techniques from broader video-language domains to further enhance performance. Our code is available at https://github.com/chidaksh/SoccerCommentary.
Authors: Ziang Chen, Liqiang Huang
Abstract: We study the approximation complexity of high-dimensional second-order elliptic PDEs with homogeneous boundary conditions on the unit hypercube, within the framework of Barron spaces. Under the assumption that the coefficients belong to suitably defined Barron spaces, we prove that the solution can be efficiently approximated by two-layer neural networks, circumventing the curse of dimensionality. Our results demonstrate the expressive power of shallow networks in capturing high-dimensional PDE solutions under appropriate structural assumptions.
Authors: Zhaorui Tan, Tan Pan, Kaizhu Huang, Weimiao Yu, Kai Yao, Chen Jiang, Qiufeng Wang, Anh Nguyen, Xin Guo, Yuan Cheng, Xi Yang
Abstract: LayerNorm is pivotal in Vision Transformers (ViTs), yet its fine-tuning dynamics under data scarcity and domain shifts remain underexplored. This paper shows that shifts in LayerNorm parameters after fine-tuning (LayerNorm shifts) are indicative of the transitions between source and target domains; its efficacy is contingent upon the degree to which the target training samples accurately represent the target domain, as quantified by our proposed Fine-tuning Shift Ratio ($FSR$). Building on this, we propose a simple yet effective rescaling mechanism using a scalar $\lambda$ that is negatively correlated to $FSR$ to align learned LayerNorm shifts with those ideal shifts achieved under fully representative data, combined with a cyclic framework that further enhances the LayerNorm fine-tuning. Extensive experiments across natural and pathological images, in both in-distribution (ID) and out-of-distribution (OOD) settings, and various target training sample regimes validate our framework. Notably, OOD tasks tend to yield lower $FSR$ and higher $\lambda$ in comparison to ID cases, especially with scarce data, indicating under-represented target training samples. Moreover, ViTFs fine-tuned on pathological data behave more like ID settings, favoring conservative LayerNorm updates. Our findings illuminate the underexplored dynamics of LayerNorm in transfer learning and provide practical strategies for LayerNorm fine-tuning.
Authors: Aswin RRV, Jacob Dineen, Divij Handa, Md Nayem Uddin, Mihir Parmar, Chitta Baral, Ben Zhou
Abstract: Recent advances in test-time scaling have led to the emergence of thinking LLMs that exhibit self-reflective behaviors and multi-step reasoning. While RL drives this self-improvement paradigm, a recent study (Gandhi et al., 2025) shows that RL alone does not truly instill these new reasoning abilities - it merely draws out behaviors already present in the base models. This raises a question: How can we train the models that don't exhibit such thinking behavior to develop it in the first place? To this end, we propose ThinkTuning, a GRPO-based interactive training approach where we augment the rollouts of a student model with the guidance from a teacher model. A simple idea from classroom practice inspires our method: a teacher poses a problem, lets the student try an answer, then gives corrective feedback -- enough to point the mind in the right direction and then show the solution. Each piece of feedback reshapes the student's thoughts, leading them to arrive at the correct solution. Similarly, we find that this type of implicit supervision through feedback from a teacher model of the same size improves the reasoning capabilities of the student model. In particular, on average, our method shows a 3.85% improvement over zero-shot baselines across benchmarks, and on MATH-500, AIME and GPQA-Diamond it shows 2.08%, 2.23% and 3.99% improvements over the vanilla-GRPO baseline. Source code is available at https://github.com/3rdAT/ThinkTuning.
Authors: Mehrshad Zandigohar, Mallesham Dasari, Gunar Schirner
Abstract: For transradial amputees, robotic prosthetic hands promise to regain the capability to perform daily living activities. To advance next-generation prosthetic hand control design, it is crucial to address current shortcomings in robustness to out of lab artifacts, and generalizability to new environments. Due to the fixed number of object to interact with in existing datasets, contrasted with the virtually infinite variety of objects encountered in the real world, current grasp models perform poorly on unseen objects, negatively affecting users' independence and quality of life. To address this: (i) we define semantic projection, the ability of a model to generalize to unseen object types and show that conventional models like YOLO, despite 80% training accuracy, drop to 15% on unseen objects. (ii) we propose Grasp-LLaVA, a Grasp Vision Language Model enabling human-like reasoning to infer the suitable grasp type estimate based on the object's physical characteristics resulting in a significant 50.2% accuracy over unseen object types compared to 36.7% accuracy of an SOTA grasp estimation model. Lastly, to bridge the performance-latency gap, we propose Hybrid Grasp Network (HGN), an edge-cloud deployment infrastructure enabling fast grasp estimation on edge and accurate cloud inference as a fail-safe, effectively expanding the latency vs. accuracy Pareto. HGN with confidence calibration (DC) enables dynamic switching between edge and cloud models, improving semantic projection accuracy by 5.6% (to 42.3%) with 3.5x speedup over the unseen object types. Over a real-world sample mix, it reaches 86% average accuracy (12.2% gain over edge-only), and 2.2x faster inference than Grasp-LLaVA alone.
Authors: Jie Li, Haoye Dong, Zhengyang Wu, Zetao Zheng, Mingrong Lin
Abstract: Next Point-of-Interest (POI) recommendation is a research hotspot in business intelligence, where users' spatial-temporal transitions and social relationships play key roles. However, most existing works model spatial and temporal transitions separately, leading to misaligned representations of the same spatial-temporal key nodes. This misalignment introduces redundant information during fusion, increasing model uncertainty and reducing interpretability. To address this issue, we propose DiMuST, a socially enhanced POI recommendation model based on disentangled representation learning over multiplex spatial-temporal transition graphs. The model employs a novel Disentangled variational multiplex graph Auto-Encoder (DAE), which first disentangles shared and private distributions using a multiplex spatial-temporal graph strategy. It then fuses the shared features via a Product of Experts (PoE) mechanism and denoises the private features through contrastive constraints. The model effectively captures the spatial-temporal transition representations of POIs while preserving the intrinsic correlation of their spatial-temporal relationships. Experiments on two challenging datasets demonstrate that our DiMuST significantly outperforms existing methods across multiple metrics.
Authors: Gianluca Bontempi
Abstract: Though intelligent agents are supposed to improve human experience (or make it more efficient), it is hard from a human perspective to grasp the ethical values which are explicitly or implicitly embedded in an agent behaviour. This is the well-known problem of alignment, which refers to the challenge of designing AI systems that align with human values, goals and preferences. This problem is particularly challenging since most human ethical considerations refer to \emph{incommensurable} (i.e. non-measurable and/or incomparable) values and criteria. Consider, for instance, a medical agent prescribing a treatment to a cancerous patient. How could it take into account (and/or weigh) incommensurable aspects like the value of a human life and the cost of the treatment? Now, the alignment between human and artificial values is possible only if we define a common space where a metric can be defined and used. This paper proposes to extend to ethics the conventional Anything2vec approach, which has been successful in plenty of similar and hard-to-quantify domains (ranging from natural language processing to recommendation systems and graph analysis). This paper proposes a way to map an automatic agent decision-making (or control law) strategy to a multivariate vector representation, which can be used to compare and assess the alignment with human values. The Ethics2Vec method is first introduced in the case of an automatic agent performing binary decision-making. Then, a vectorisation of an automatic control law (like in the case of a self-driving car) is discussed to show how the approach can be extended to automatic control settings.
Authors: Markus Fritzsche, Elliot Gestrin, Jendrik Seipp
Abstract: While transformers excel in many settings, their application in the field of automated planning is limited. Prior work like PlanGPT, a state-of-the-art decoder-only transformer, struggles with extrapolation from easy to hard planning problems. This in turn stems from problem symmetries: planning tasks can be represented with arbitrary variable names that carry no meaning beyond being identifiers. This causes a combinatorial explosion of equivalent representations that pure transformers cannot efficiently learn from. We propose a novel contrastive learning objective to make transformers symmetry-aware and thereby compensate for their lack of inductive bias. Combining this with architectural improvements, we show that transformers can be efficiently trained for either plan-generation or heuristic-prediction. Our results across multiple planning domains demonstrate that our symmetry-aware training effectively and efficiently addresses the limitations of PlanGPT.
Authors: Mohammed Salah, Numan Saeed, Davor Svetinovic, Stefano Sfarra, Mohammed Omar, Yusra Abdulrahman
Abstract: Active Infrared thermography (AIRT) is a widely adopted non-destructive testing (NDT) technique for detecting subsurface anomalies in industrial components. Due to the high dimensionality of AIRT data, current approaches employ non-linear autoencoders (AEs) for dimensionality reduction. However, the latent space learned by AIRT AEs lacks structure, limiting their effectiveness in downstream defect characterization tasks. To address this limitation, this paper proposes a principal component analysis guided (PCA-guided) autoencoding framework for structured dimensionality reduction to capture intricate, non-linear features in thermographic signals while enforcing a structured latent space. A novel loss function, PCA distillation loss, is introduced to guide AIRT AEs to align the latent representation with structured PCA components while capturing the intricate, non-linear patterns in thermographic signals. To evaluate the utility of the learned, structured latent space, we propose a neural network-based evaluation metric that assesses its suitability for defect characterization. Experimental results show that the proposed PCA-guided AE outperforms state-of-the-art dimensionality reduction methods on PVC, CFRP, and PLA samples in terms of contrast, signal-to-noise ratio (SNR), and neural network-based metrics.
Authors: Cheng Li, Pengfei Danga, Yuehui Xiana, Yumei Zhou, Bofeng Shi, Xiangdong Ding, Jun Suna, Dezhen Xue
Abstract: The design of shape memory alloys (SMAs) with high transformation temperatures and large mechanical work output remains a longstanding challenge in functional materials engineering. Here, we introduce a data-driven framework based on generative adversarial network (GAN) inversion for the inverse design of high-performance SMAs. By coupling a pretrained GAN with a property prediction model, we perform gradient-based latent space optimization to directly generate candidate alloy compositions and processing parameters that satisfy user-defined property targets. The framework is experimentally validated through the synthesis and characterization of five NiTi-based SMAs. Among them, the Ni$_{49.8}$Ti$_{26.4}$Hf$_{18.6}$Zr$_{5.2}$ alloy achieves a high transformation temperature of 404 $^\circ$C, a large mechanical work output of 9.9 J/cm$^3$, a transformation enthalpy of 43 J/g , and a thermal hysteresis of 29 {\deg}C, outperforming existing NiTi alloys. The enhanced performance is attributed to a pronounced transformation volume change and a finely dispersed of Ti$_2$Ni-type precipitates, enabled by sluggish Zr and Hf diffusion, and semi-coherent interfaces with localized strain fields. This study demonstrates that GAN inversion offers an efficient and generalizable route for the property-targeted discovery of complex alloys.
Authors: Tao Tang, Chengxu Yang
Abstract: The core role of medical images in disease diagnosis makes their quality directly affect the accuracy of clinical judgment. However, due to factors such as low-dose scanning, equipment limitations and imaging artifacts, medical images are often accompanied by non-uniform noise interference, which seriously affects structure recognition and lesion detection. This paper proposes a medical image adaptive denoising model (MI-ND) that integrates multi-scale convolutional and Transformer architecture, introduces a noise level estimator (NLE) and a noise adaptive attention module (NAAB), and realizes channel-spatial attention regulation and cross-modal feature fusion driven by noise perception. Systematic testing is carried out on multimodal public datasets. Experiments show that this method significantly outperforms the comparative methods in image quality indicators such as PSNR, SSIM, and LPIPS, and improves the F1 score and ROC-AUC in downstream diagnostic tasks, showing strong prac-tical value and promotional potential. The model has outstanding benefits in structural recovery, diagnostic sensitivity, and cross-modal robustness, and provides an effective solution for medical image enhancement and AI-assisted diagnosis and treatment.
Authors: Ke Ma, Jun Long, Hongxiao Fei, Liujie Hua, Yueyi Luo
Abstract: Pre-trained Vision-Language Models (VLMs) face a significant adaptation gap when applied to Zero-Shot Anomaly Detection (ZSAD), stemming from their lack of local inductive biases for dense prediction and their reliance on inflexible feature fusion paradigms. We address these limitations through an Architectural Co-Design framework that jointly refines feature representation and cross-modal fusion. Our method integrates a parameter-efficient Convolutional Low-Rank Adaptation (Conv-LoRA) adapter to inject local inductive biases for fine-grained representation, and introduces a Dynamic Fusion Gateway (DFG) that leverages visual context to adaptively modulate text prompts, enabling a powerful bidirectional fusion. Extensive experiments on diverse industrial and medical benchmarks demonstrate superior accuracy and robustness, validating that this synergistic co-design is critical for robustly adapting foundation models to dense perception tasks.
Authors: Vishwas M. Shetty, Jiusi Zheng, Abeer Alwan
Abstract: Speaker Verification (SV) systems trained on adults speech often underperform on children's SV due to the acoustic mismatch, and limited children speech data makes fine-tuning not very effective. In this paper, we propose an innovative framework, a Gated Linear Unit adapter with Iterative Fine-Tuning (G-IFT), to enhance knowledge transfer efficiency between the high-resource adults speech domain and the low-resource children's speech domain. In this framework, a Gated Linear Unit adapter is first inserted between the pre-trained speaker embedding model and the classifier. Then the classifier, adapter, and pre-trained speaker embedding model are optimized sequentially in an iterative way. This framework is agnostic to the type of the underlying architecture of the SV system. Our experiments on ECAPA-TDNN, ResNet, and X-vector architectures using the OGI and MyST datasets demonstrate that the G-IFT framework yields consistent reductions in Equal Error Rates compared to baseline methods.
Authors: Danil Gusak, Nikita Sukhorukov, Evgeny Frolov
Abstract: In modern recommender systems, experimental settings typically include filtering out cold users and items based on a minimum interaction threshold. However, these thresholds are often chosen arbitrarily and vary widely across studies, leading to inconsistencies that can significantly affect the comparability and reliability of evaluation results. In this paper, we systematically explore the cold-start boundary by examining the criteria used to determine whether a user or an item should be considered cold. Our experiments incrementally vary the number of interactions for different items during training, and gradually update the length of user interaction histories during inference. We investigate the thresholds across several widely used datasets, commonly represented in recent papers from top-tier conferences, and on multiple established recommender baselines. Our findings show that inconsistent selection of cold-start thresholds can either result in the unnecessary removal of valuable data or lead to the misclassification of cold instances as warm, introducing more noise into the system.
Authors: Bin Cao, Sipeng Zheng, Ye Wang, Lujie Xia, Qianshan Wei, Qin Jin, Jing Liu, Zongqing Lu
Abstract: Human motion generation has emerged as a critical technology with transformative potential for real-world applications. However, existing vision-language-motion models (VLMMs) face significant limitations that hinder their practical deployment. We identify controllability as a main bottleneck, manifesting in five key aspects: inadequate response to diverse human commands, limited pose initialization capabilities, poor performance on long-term sequences, insufficient handling of unseen scenarios, and lack of fine-grained control over individual body parts. To overcome these limitations, we present Being-M0.5, the first real-time, controllable VLMM that achieves state-of-the-art performance across multiple motion generation tasks. Our approach is built upon HuMo100M, the largest and most comprehensive human motion dataset to date, comprising over 5 million self-collected motion sequences, 100 million multi-task instructional instances, and detailed part-level annotations that address a critical gap in existing datasets. We introduce a novel part-aware residual quantization technique for motion tokenization that enables precise, granular control over individual body parts during generation. Extensive experimental validation demonstrates Being-M0.5's superior performance across diverse motion benchmarks, while comprehensive efficiency analysis confirms its real-time capabilities. Our contributions include design insights and detailed computational analysis to guide future development of practical motion generators. We believe that HuMo100M and Being-M0.5 represent significant advances that will accelerate the adoption of motion generation technologies in real-world applications. The project page is available at https://beingbeyond.github.io/Being-M0.5.
Authors: Holli Sargeant, Mackenzie Jorgensen, Arina Shah, Adrian Weller, Umang Bhatt
Abstract: Uncertainty in artificial intelligence (AI) predictions poses urgent legal and ethical challenges for AI-assisted decision-making. We examine two algorithmic interventions that act as guardrails for human-AI collaboration: selective abstention, which withholds high-uncertainty predictions from human decision-makers, and selective friction, which delivers those predictions together with salient warnings or disclosures that slow the decision process. Research has shown that selective abstention based on uncertainty can inadvertently exacerbate disparities and disadvantage under-represented groups that disproportionately receive uncertain predictions. In this paper, we provide the first integrated socio-technical and legal analysis of uncertainty-based algorithmic interventions. Through two case studies, AI-assisted consumer credit decisions and AI-assisted content moderation, we demonstrate how the seemingly neutral use of uncertainty thresholds can trigger discriminatory impacts. We argue that, although both interventions pose risks of unlawful discrimination under UK law, selective frictions offer a promising pathway toward fairer and more accountable AI-assisted decision-making by preserving transparency and encouraging more cautious human judgment.
Authors: Samaneh Mohammadi, Vasileios Tsouvalas, Iraklis Symeonidis, Ali Balador, Tanir Ozcelebi, Francesco Flammini, Nirvana Meratnia
Abstract: Federated unlearning (FU) algorithms allow clients in federated settings to exercise their ''right to be forgotten'' by removing the influence of their data from a collaboratively trained model. Existing FU methods maintain data privacy by performing unlearning locally on the client-side and sending targeted updates to the server without exposing forgotten data; yet they often rely on server-side cooperation, revealing the client's intent and identity without enforcement guarantees - compromising autonomy and unlearning privacy. In this work, we propose EFU (Enforced Federated Unlearning), a cryptographically enforced FU framework that enables clients to initiate unlearning while concealing its occurrence from the server. Specifically, EFU leverages functional encryption to bind encrypted updates to specific aggregation functions, ensuring the server can neither perform unauthorized computations nor detect or skip unlearning requests. To further mask behavioral and parameter shifts in the aggregated model, we incorporate auxiliary unlearning losses based on adversarial examples and parameter importance regularization. Extensive experiments show that EFU achieves near-random accuracy on forgotten data while maintaining performance comparable to full retraining across datasets and neural architectures - all while concealing unlearning intent from the server. Furthermore, we demonstrate that EFU is agnostic to the underlying unlearning algorithm, enabling secure, function-hiding, and verifiable unlearning for any client-side FU mechanism that issues targeted updates.
Authors: Juan-Pablo Ortega, Florian Rossmannek
Abstract: This work advances the theoretical foundations of reservoir computing (RC) by providing a unified treatment of fading memory and the echo state property (ESP) in both deterministic and stochastic settings. We investigate state-space systems, a central model class in time series learning, and establish that fading memory and solution stability hold generically -- even in the absence of the ESP -- offering a robust explanation for the empirical success of RC models without strict contractivity conditions. In the stochastic case, we critically assess stochastic echo states, proposing a novel distributional perspective rooted in attractor dynamics on the space of probability distributions, which leads to a rich and coherent theory. Our results extend and generalize previous work on non-autonomous dynamical systems, offering new insights into causality, stability, and memory in RC models. This lays the groundwork for reliable generative modeling of temporal data in both deterministic and stochastic regimes.
Authors: Olivier Jeunen
Abstract: Off-policy estimation (OPE) methods enable unbiased offline evaluation of recommender systems, directly estimating the online reward some target policy would have obtained, from offline data and with statistical guarantees. The theoretical elegance of the framework combined with practical successes have led to a surge of interest, with many competing estimators now available to practitioners and researchers. Among these, Doubly Robust methods provide a prominent strategy to combine value- and policy-based estimators. In this work, we take an alternative perspective to combine a set of OPE estimators and their associated confidence intervals into a single, more accurate estimate. Our approach leverages a correlated fixed-effects meta-analysis framework, explicitly accounting for dependencies among estimators that arise due to shared data. This yields a best linear unbiased estimate (BLUE) of the target policy's value, along with an appropriately conservative confidence interval that reflects inter-estimator correlation. We validate our method on both simulated and real-world data, demonstrating improved statistical efficiency over existing individual estimators.
Authors: Jakub Binda, Valentina Paneta, Vasileios Eleftheriadis, Hongkyou Chung, Panagiotis Papadimitroulas, Neo Christopher Chung
Abstract: Generative AI holds great potentials to automate and enhance data synthesis in nuclear medicine. However, the high-stakes nature of biomedical imaging necessitates robust mechanisms to detect and manage unexpected or erroneous model behavior. We introduce development and implementation of a hybrid anomaly detection framework to safeguard GenAI models in BIOEMTECH's eyes(TM) systems. Two applications are demonstrated: Pose2Xray, which generates synthetic X-rays from photographic mouse images, and DosimetrEYE, which estimates 3D radiation dose maps from 2D SPECT/CT scans. In both cases, our outlier detection (OD) enhances reliability, reduces manual oversight, and supports real-time quality control. This approach strengthens the industrial viability of GenAI in preclinical settings by increasing robustness, scalability, and regulatory compliance.
Authors: Bogdan Butyrin, Artemy Rubtsov, Alexey Naumov, Vladimir Ulyanov, Sergey Samsonov
Abstract: In this paper, we establish non-asymptotic bounds for accuracy of normal approximation for linear two-timescale stochastic approximation (TTSA) algorithms driven by martingale difference or Markov noise. Focusing on both the last iterate and Polyak-Ruppert averaging regimes, we derive bounds for normal approximation in terms of the convex distance between probability distributions. Our analysis reveals a non-trivial interaction between the fast and slow timescales: the normal approximation rate for the last iterate improves as the timescale separation increases, while it decreases in the Polyak-Ruppert averaged setting. We also provide the high-order moment bounds for the error of linear TTSA algorithm, which may be of independent interest.
Authors: Chen Shen, Wanqing Zhang, Kehan Li, Erwen Huang, Haitao Bi, Aiying Fan, Yiwen Shen, Hongmei Dong, Ji Zhang, Yuming Shao, Zengjia Liu, Xinshe Liu, Tao Li, Chunxia Yan, Shuanliang Fan, Di Wu, Jianhua Ma, Bin Cong, Zhenyuan Wang, Chunfeng Lian
Abstract: Forensic cause-of-death determination faces systemic challenges, including workforce shortages and diagnostic variability, particularly in high-volume systems like China's medicolegal infrastructure. We introduce FEAT (ForEnsic AgenT), a multi-agent AI framework that automates and standardizes death investigations through a domain-adapted large language model. FEAT's application-oriented architecture integrates: (i) a central Planner for task decomposition, (ii) specialized Local Solvers for evidence analysis, (iii) a Memory & Reflection module for iterative refinement, and (iv) a Global Solver for conclusion synthesis. The system employs tool-augmented reasoning, hierarchical retrieval-augmented generation, forensic-tuned LLMs, and human-in-the-loop feedback to ensure legal and medical validity. In evaluations across diverse Chinese case cohorts, FEAT outperformed state-of-the-art AI systems in both long-form autopsy analyses and concise cause-of-death conclusions. It demonstrated robust generalization across six geographic regions and achieved high expert concordance in blinded validations. Senior pathologists validated FEAT's outputs as comparable to those of human experts, with improved detection of subtle evidentiary nuances. To our knowledge, FEAT is the first LLM-based AI agent system dedicated to forensic medicine, offering scalable, consistent death certification while maintaining expert-level rigor. By integrating AI efficiency with human oversight, this work could advance equitable access to reliable medicolegal services while addressing critical capacity constraints in forensic systems.
Authors: Dongxu Li, Kai Yuan, Jianhao Huang, Chuan Huang, Xiaoqi Qin, Shuguang Cui, Ping Zhang
Abstract: Semantic communications (SemComs) have emerged as a promising paradigm for joint data and task-oriented transmissions, combining the demands for both the bit-accurate delivery and end-to-end (E2E) distortion minimization. However, current joint source-channel coding (JSCC) in SemComs is not compatible with the existing communication systems and cannot adapt to the variations of the sources or the channels, while separate source-channel coding (SSCC) is suboptimal in the finite blocklength regime. To address these issues, we propose an adaptive source-channel coding (ASCC) scheme for SemComs over parallel Gaussian channels, where the deep neural network (DNN)-based semantic source coding and conventional digital channel coding are separately deployed and adaptively designed. To enable efficient adaptation between the source and channel coding, we first approximate the E2E data and semantic distortions as functions of source coding rate and bit error ratio (BER) via logistic regression, where BER is further modeled as functions of signal-to-noise ratio (SNR) and channel coding rate. Then, we formulate the weighted sum E2E distortion minimization problem for joint source-channel coding rate and power allocation over parallel channels, which is solved by the successive convex approximation. Finally, simulation results demonstrate that the proposed ASCC scheme outperforms typical deep JSCC and SSCC schemes for both the single- and parallel-channel scenarios while maintaining full compatibility with practical digital systems.
Authors: Leonardo V. Santoro, Victor M. Panaretos
Abstract: We propose a novel kernel-based nonparametric two-sample test, employing the combined use of kernel mean and kernel covariance embedding. Our test builds on recent results showing how such combined embeddings map distinct probability measures to mutually singular Gaussian measures on the kernel's RKHS. Leveraging this result, we construct a test statistic based on the relative entropy between the Gaussian embeddings, i.e.\ the likelihood ratio. The likelihood ratio is specifically tailored to detect equality versus singularity of two Gaussians, and satisfies a ``$0/\infty$" law, in that it vanishes under the null and diverges under the alternative. To implement the test in finite samples, we introduce a regularised version, calibrated by way of permutation. We prove consistency, establish uniform power guarantees under mild conditions, and discuss how our framework unifies and extends prior approaches based on spectrally regularized MMD. Empirical results on synthetic and real data demonstrate remarkable gains in power compared to state-of-the-art methods, particularly in high-dimensional and weak-signal regimes.
Authors: Pierre Perrault
Abstract: This paper presents an improved exponential tail bound for Beta distributions, refining a result in [15]. This improvement is achieved by interpreting their bound as a regular Kullback-Leibler (KL) divergence one, while introducing a specific perturbation $\eta$ that shifts the mean of the Beta distribution closer to zero within the KL bound. Our contribution is to show that a larger perturbation can be chosen, thereby tightening the bound. We then extend this result from the Beta distribution to Dirichlet distributions and Dirichlet processes (DPs).
Authors: Birgit Hillebrecht, Benjamin Unger
Abstract: Rigorous error estimation is a fundamental topic in numerical analysis. With the increasing use of physics-informed neural networks (PINNs) for solving partial differential equations, several approaches have been developed to quantify the associated prediction error. In this work, we build upon a semigroup-based framework previously introduced by the authors for estimating the PINN error. While this estimator has so far been limited to academic examples - due to the need to compute quantities related to input-to-state stability - we extend its applicability to a significantly broader class of problems. This is accomplished by modifying the error bound and proposing numerical strategies to approximate the required stability parameters. The extended framework enables the certification of PINN predictions in more realistic scenarios, as demonstrated by a numerical study of Stokes flow around a cylinder.
Authors: Hengyu Liu, Yushuai Li, Minghe Yu, Tiancheng Zhang, Ge Yu, Torben Bach Pedersen, Kristian Torp, Christian S. Jensen, Tianyi Li
Abstract: Intelligent Tutoring Systems (ITS), such as Massive Open Online Courses, offer new opportunities for human learning. At the core of such systems, knowledge tracing (KT) predicts students' future performance by analyzing their historical learning activities, enabling an accurate evaluation of students' knowledge states over time. We show that existing KT methods often encounter correlation conflicts when analyzing the relationships between historical learning sequences and future performance. To address such conflicts, we propose to extract so-called Follow-up Performance Trends (FPTs) from historical ITS data and to incorporate them into KT. We propose a method called Forward-Looking Knowledge Tracing (FINER) that combines historical learning sequences with FPTs to enhance student performance prediction accuracy. FINER constructs learning patterns that facilitate the retrieval of FPTs from historical ITS data in linear time; FINER includes a novel similarity-aware attention mechanism that aggregates FPTs based on both frequency and contextual similarity; and FINER offers means of combining FPTs and historical learning sequences to enable more accurate prediction of student future performance. Experiments on six real-world datasets show that FINER can outperform ten state-of-the-art KT methods, increasing accuracy by 8.74% to 84.85%.
Authors: Roopkatha Banerjee, Sampath Koti, Gyanendra Singh, Anirban Chakraborty, Gurunath Gurrala, Bhushan Jagyasi, Yogesh Simmhan
Abstract: Real-time monitoring of power consumption in cities and micro-grids through the Internet of Things (IoT) can help forecast future demand and optimize grid operations. But moving all consumer-level usage data to the cloud for predictions and analysis at fine time scales can expose activity patterns. Federated Learning~(FL) is a privacy-sensitive collaborative DNN training approach that retains data on edge devices, trains the models on private data locally, and aggregates the local models in the cloud. But key challenges exist: (i) clients can have non-independently identically distributed~(non-IID) data, and (ii) the learning should be computationally cheap while scaling to 1000s of (unseen) clients. In this paper, we develop and evaluate several optimizations to FL training across edge and cloud for time-series demand forecasting in micro-grids and city-scale utilities using DNNs to achieve a high prediction accuracy while minimizing the training cost. We showcase the benefit of using exponentially weighted loss while training and show that it further improves the prediction of the final model. Finally, we evaluate these strategies by validating over 1000s of clients for three states in the US from the OpenEIA corpus, and performing FL both in a pseudo-distributed setting and a Pi edge cluster. The results highlight the benefits of the proposed methods over baselines like ARIMA and DNNs trained for individual consumers, which are not scalable.
Authors: Thusitha Dayaratne, Ngoc Duy Pham, Viet Vo, Shangqi Lai, Sharif Abuadbba, Hajime Suzuki, Xingliang Yuan, Carsten Rudolph
Abstract: The introduction of 5G and the Open Radio Access Network (O-RAN) architecture has enabled more flexible and intelligent network deployments. However, the increased complexity and openness of these architectures also introduce novel security challenges, such as data manipulation attacks on the semi-standardised Shared Data Layer (SDL) within the O-RAN platform through malicious xApps. In particular, malicious xApps can exploit this vulnerability by introducing subtle Unicode-wise alterations (hypoglyphs) into the data that are being used by traditional machine learning (ML)-based anomaly detection methods. These Unicode-wise manipulations can potentially bypass detection and cause failures in anomaly detection systems based on traditional ML, such as AutoEncoders, which are unable to process hypoglyphed data without crashing. We investigate the use of Large Language Models (LLMs) for anomaly detection within the O-RAN architecture to address this challenge. We demonstrate that LLM-based xApps maintain robust operational performance and are capable of processing manipulated messages without crashing. While initial detection accuracy requires further improvements, our results highlight the robustness of LLMs to adversarial attacks such as hypoglyphs in input data. There is potential to use their adaptability through prompt engineering to further improve the accuracy, although this requires further research. Additionally, we show that LLMs achieve low detection latency (under 0.07 seconds), making them suitable for Near-Real-Time (Near-RT) RIC deployments.
Authors: Hao Peng, Yuanyuan Zhang, Steve Jiang, Robert Timmerman, John Minna
Abstract: Radiation response in cancer is shaped by complex, patient specific biology, yet current treatment strategies often rely on uniform dose prescriptions without accounting for tumor heterogeneity. In this study, we introduce a meta learning framework for one-shot prediction of radiosensitivity measured by SF2 using cell line level gene expression data. Unlike the widely used Radiosensitivity Index RSI a rank-based linear model trained on a fixed 10-gene signature, our proposed meta-learned model allows the importance of each gene to vary by sample through fine tuning. This flexibility addresses key limitations of static models like RSI, which assume uniform gene contributions across tumor types and discard expression magnitude and gene gene interactions. Our results show that meta learning offers robust generalization to unseen samples and performs well in tumor subgroups with high radiosensitivity variability, such as adenocarcinoma and large cell carcinoma. By learning transferable structure across tasks while preserving sample specific adaptability, our approach enables rapid adaptation to individual samples, improving predictive accuracy across diverse tumor subtypes while uncovering context dependent patterns of gene influence that may inform personalized therapy.
Authors: Ziad Al-Haj Hemidi, Eytan Kats, Mattias P. Heinrich
Abstract: Accelerating Magnetic Resonance Imaging (MRI) reduces scan time but often degrades image quality. While Implicit Neural Representations (INRs) show promise for MRI reconstruction, they struggle at high acceleration factors due to weak prior constraints, leading to structural loss and aliasing artefacts. To address this, we propose PrIINeR, an INR-based MRI reconstruction method that integrates prior knowledge from pre-trained deep learning models into the INR framework. By combining population-level knowledge with instance-based optimization and enforcing dual data consistency, PrIINeR aligns both with the acquired k-space data and the prior-informed reconstruction. Evaluated on the NYU fastMRI dataset, our method not only outperforms state-of-the-art INR-based approaches but also improves upon several learning-based state-of-the-art methods, significantly improving structural preservation and fidelity while effectively removing aliasing artefacts.PrIINeR bridges deep learning and INR-based techniques, offering a more reliable solution for high-quality, accelerated MRI reconstruction. The code is publicly available on https://github.com/multimodallearning/PrIINeR.
Authors: Weitai Kang, Weiming Zhuang, Zhizhong Li, Yan Yan, Lingjuan Lyu
Abstract: Fine-grained multimodal capability in Multimodal Large Language Models (MLLMs) has emerged as a critical research direction, particularly for tackling the visual grounding (VG) problem. Despite the strong performance achieved by existing approaches, they often employ disparate design choices when fine-tuning MLLMs for VG, lacking systematic verification to support these designs. To bridge this gap, this paper presents a comprehensive study of various design choices that impact the VG performance of MLLMs. We conduct our analysis using LLaVA-1.5, which has been widely adopted in prior empirical studies of MLLMs. While more recent models exist, we follow this convention to ensure our findings remain broadly applicable and extendable to other architectures. We cover two key aspects: (1) exploring different visual grounding paradigms in MLLMs, identifying the most effective design, and providing our insights; and (2) conducting ablation studies on the design of grounding data to optimize MLLMs' fine-tuning for the VG task. Finally, our findings contribute to a stronger MLLM for VG, achieving improvements of +5.6% / +6.9% / +7.0% on RefCOCO/+/g over the LLaVA-1.5.
Authors: Md Rezwanul Haque, Md. Milon Islam, S M Taslim Uddin Raju, Hamdi Altaheri, Lobna Nassar, Fakhri Karray
Abstract: Depression is a major mental health condition that severely impacts the emotional and physical well-being of individuals. The simple nature of data collection from social media platforms has attracted significant interest in properly utilizing this information for mental health research. A Multimodal Depression Detection Network (MDD-Net), utilizing acoustic and visual data obtained from social media networks, is proposed in this work where mutual transformers are exploited to efficiently extract and fuse multimodal features for efficient depression detection. The MDD-Net consists of four core modules: an acoustic feature extraction module for retrieving relevant acoustic attributes, a visual feature extraction module for extracting significant high-level patterns, a mutual transformer for computing the correlations among the generated features and fusing these features from multiple modalities, and a detection layer for detecting depression using the fused feature representations. The extensive experiments are performed using the multimodal D-Vlog dataset, and the findings reveal that the developed multimodal depression detection network surpasses the state-of-the-art by up to 17.37% for F1-Score, demonstrating the greater performance of the proposed system. The source code is accessible at https://github.com/rezwanh001/Multimodal-Depression-Detection.
URLs: https://github.com/rezwanh001/Multimodal-Depression-Detection.
Authors: Lukas Gehring, Benjamin Paa{\ss}en
Abstract: Recent advancements in Large Language Models (LLMs) and their increased accessibility have made it easier than ever for students to automatically generate texts, posing new challenges for educational institutions. To enforce norms of academic integrity and ensure students' learning, learning analytics methods to automatically detect LLM-generated text appear increasingly appealing. This paper benchmarks the performance of different state-of-the-art detectors in educational contexts, introducing a novel dataset, called Generative Essay Detection in Education (GEDE), containing over 900 student-written essays and over 12,500 LLM-generated essays from various domains. To capture the diversity of LLM usage practices in generating text, we propose the concept of contribution levels, representing students' contribution to a given assignment. These levels range from purely human-written texts, to slightly LLM-improved versions, to fully LLM-generated texts, and finally to active attacks on the detector by "humanizing" generated texts. We show that most detectors struggle to accurately classify texts of intermediate student contribution levels, like LLM-improved human-written texts. Detectors are particularly likely to produce false positives, which is problematic in educational settings where false suspicions can severely impact students' lives. Our dataset, code, and additional supplementary materials are publicly available at https://github.com/lukasgehring/Assessing-LLM-Text-Detection-in-Educational-Contexts.
URLs: https://github.com/lukasgehring/Assessing-LLM-Text-Detection-in-Educational-Contexts.
Authors: Adrian Baule
Abstract: Score-based diffusion models generate samples from a complex underlying data distribution by time-reversal of a diffusion process and represent the state-of-the-art in many generative AI applications such as artificial image synthesis. Here, I show how a generative diffusion model can be implemented based on an underlying active particle process with finite correlation time. In contrast to previous approaches that use a score function acting on the velocity coordinate of the active particle, time reversal is here achieved by imposing an effective time-dependent potential on the position coordinate only. The effective potential is valid to first order in the persistence time and leads to a force field that is fully determined by the standard score function and its derivatives up to 2nd order. Numerical experiments for artificial data distributions confirm the validity of the effective potential.
Authors: Mandira Sawkar, Samay U. Shetty, Deepak Pandita, Tharindu Cyril Weerasooriya, Christopher M. Homan
Abstract: The Learning With Disagreements (LeWiDi) 2025 shared task is to model annotator disagreement through soft label distribution prediction and perspectivist evaluation, modeling annotators. We adapt DisCo (Distribution from Context), a neural architecture that jointly models item-level and annotator-level label distributions, and present detailed analysis and improvements. In this paper, we extend the DisCo by incorporating annotator metadata, enhancing input representations, and modifying the loss functions to capture disagreement patterns better. Through extensive experiments, we demonstrate substantial improvements in both soft and perspectivist evaluation metrics across three datasets. We also conduct in-depth error and calibration analyses, highlighting the conditions under which improvements occur. Our findings underscore the value of disagreement-aware modeling and offer insights into how system components interact with the complexity of human-annotated data.
Authors: Yan Wang, Da-Wei Zhou, Han-Jia Ye
Abstract: Class-Incremental Learning (CIL) requires a learning system to continually learn new classes without forgetting. Existing pre-trained model-based CIL methods often freeze the pre-trained network and adapt to incremental tasks using additional lightweight modules such as adapters. However, incorrect module selection during inference hurts performance, and task-specific modules often overlook shared general knowledge, leading to errors on distinguishing between similar classes across tasks. To address the aforementioned challenges, we propose integrating Task-Specific and Universal Adapters (TUNA) in this paper. Specifically, we train task-specific adapters to capture the most crucial features relevant to their respective tasks and introduce an entropy-based selection mechanism to choose the most suitable adapter. Furthermore, we leverage an adapter fusion strategy to construct a universal adapter, which encodes the most discriminative features shared across tasks. We combine task-specific and universal adapter predictions to harness both specialized and general knowledge during inference. Extensive experiments on various benchmark datasets demonstrate the state-of-the-art performance of our approach. Code is available at: https://github.com/LAMDA-CL/ICCV2025-TUNA
Authors: Amirhossein Taherpour, Abbas Taherpour, Tamer Khattab
Abstract: We propose a joint learning framework for Byzantine-resilient spectrum sensing and secure intelligent reflecting surface (IRS)--assisted opportunistic access under channel state information (CSI) uncertainty. The sensing stage performs logit-domain Bayesian updates with trimmed aggregation and attention-weighted consensus, and the base station (BS) fuses network beliefs with a conservative minimum rule, preserving detection accuracy under a bounded number of Byzantine users. Conditioned on the sensing outcome, we pose downlink design as sum mean-squared error (MSE) minimization under transmit-power and signal-leakage constraints and jointly optimize the BS precoder, IRS phase shifts, and user equalizers. With partial (or known) CSI, we develop an augmented-Lagrangian alternating algorithm with projected updates and provide provable sublinear convergence, with accelerated rates under mild local curvature. With unknown CSI, we perform constrained Bayesian optimization (BO) in a geometry-aware low-dimensional latent space using Gaussian process (GP) surrogates; we prove regret bounds for a constrained upper confidence bound (UCB) variant of the BO module, and demonstrate strong empirical performance of the implemented procedure. Simulations across diverse network conditions show higher detection probability at fixed false-alarm rate under adversarial attacks, large reductions in sum MSE for honest users, strong suppression of eavesdropper signal power, and fast convergence. The framework offers a practical path to secure opportunistic communication that adapts to CSI availability while coherently coordinating sensing and transmission through joint learning.
Authors: Zhuohao Yu, Xingru Jiang, Weizheng Gu, Yidong Wang, Shikun Zhang, Wei Ye
Abstract: Watermarking LLM-generated text is critical for content attribution and misinformation prevention. However, existing methods compromise text quality, require white-box model access and logit manipulation. These limitations exclude API-based models and multilingual scenarios. We propose SAEMark, a general framework for post-hoc multi-bit watermarking that embeds personalized messages solely via inference-time, feature-based rejection sampling without altering model logits or requiring training. Our approach operates on deterministic features extracted from generated text, selecting outputs whose feature statistics align with key-derived targets. This framework naturally generalizes across languages and domains while preserving text quality through sampling LLM outputs instead of modifying. We provide theoretical guarantees relating watermark success probability and compute budget that hold for any suitable feature extractor. Empirically, we demonstrate the framework's effectiveness using Sparse Autoencoders (SAEs), achieving superior detection accuracy and text quality. Experiments across 4 datasets show SAEMark's consistent performance, with 99.7% F1 on English and strong multi-bit detection accuracy. SAEMark establishes a new paradigm for scalable watermarking that works out-of-the-box with closed-source LLMs while enabling content attribution.
Authors: Zilong Zhao, Robert Birke, Aditya Kunar, Lydia Y. Chen
Abstract: Generative Adversarial Networks (GANs) are typically trained to synthesize data, from images and more recently tabular data, under the assumption of directly accessible training data. Recently, federated learning (FL) is an emerging paradigm that features decentralized learning on client's local data with a privacy-preserving capability. And, while learning GANs to synthesize images on FL systems has just been demonstrated, it is unknown if GANs for tabular data can be learned from decentralized data sources. Moreover, it remains unclear which distributed architecture suits them best. Different from image GANs, state-of-the-art tabular GANs require prior knowledge on the data distribution of each (discrete and continuous) column to agree on a common encoding -- risking privacy guarantees. In this paper, we propose Fed-TGAN, the first Federated learning framework for Tabular GANs. To effectively learn a complex tabular GAN on non-identical participants, Fed-TGAN designs two novel features: (i) a privacy-preserving multi-source feature encoding for model initialization; and (ii) table similarity aware weighting strategies to aggregate local models for countering data skew. We extensively evaluate the proposed Fed-TGAN against variants of decentralized learning architectures on four widely used datasets. Results show that Fed-TGAN accelerates training time per epoch up to 200% compared to the alternative architectures, for both IID and Non-IID data. Overall, Fed-TGAN not only stabilizes the training loss, but also achieves better similarity between generated and original data. Our code is released at https://github.com/zhao-zilong/Fed-TGAN.
Authors: S. Fatemeh Seyyedsalehi, Mahdieh Soleymani, Hamid R. Rabiee
Abstract: We propose a novel interpretation technique to explain the behavior of structured output models, which learn mappings between an input vector to a set of output variables simultaneously. Because of the complex relationship between the computational path of output variables in structured models, a feature can affect the value of output through other ones. We focus on one of the outputs as the target and try to find the most important features utilized by the structured model to decide on the target in each locality of the input space. In this paper, we assume an arbitrary structured output model is available as a black box and argue how considering the correlations between output variables can improve the explanation performance. The goal is to train a function as an interpreter for the target output variable over the input space. We introduce an energy-based training process for the interpreter function, which effectively considers the structural information incorporated into the model to be explained. The effectiveness of the proposed method is confirmed using a variety of simulated and real data sets.
Authors: Farha A. Khan, Tanmay Chakraborty, J\"org P. Dietrich, Christian Wirth
Abstract: State-of-the-art multi-objective optimization often assumes a known utility function, learns it interactively, or computes the full Pareto front-each requiring costly expert input.~Real-world problems, however, involve implicit preferences that are hard to formalize. To reduce expert involvement, we propose an offline, interpretable utility learning method that uses expert knowledge, historical examples, and coarse information about the utility space to reduce sample requirements. We model uncertainty via a full Bayesian posterior and propagate it throughout the optimization process. Our method outperforms standard Gaussian processes and BOPE across four domains, showing strong performance even with biased samples, as encountered in the real-world, and limited expert input.
Authors: Mikael M{\o}ller H{\o}gsgaard, Kasper Green Larsen, Martin Ritzert
Abstract: AdaBoost is a classic boosting algorithm for combining multiple inaccurate classifiers produced by a weak learner, to produce a strong learner with arbitrarily high accuracy when given enough training data. Determining the optimal number of samples necessary to obtain a given accuracy of the strong learner, is a basic learning theoretic question. Larsen and Ritzert (NeurIPS'22) recently presented the first provably optimal weak-to-strong learner. However, their algorithm is somewhat complicated and it remains an intriguing question whether the prototypical boosting algorithm AdaBoost also makes optimal use of training samples. In this work, we answer this question in the negative. Concretely, we show that the sample complexity of AdaBoost, and other classic variations thereof, are sub-optimal by at least one logarithmic factor in the desired accuracy of the strong learner.
Authors: Xuefeng Liu, Takuma Yoneda, Chaoqi Wang, Matthew R. Walter, Yuxin Chen
Abstract: Reinforcement learning (RL) has made significant strides in various complex domains. However, identifying an effective policy via RL often necessitates extensive exploration. Imitation learning aims to mitigate this issue by using expert demonstrations to guide exploration. In real-world scenarios, one often has access to multiple suboptimal black-box experts, rather than a single optimal oracle. These experts do not universally outperform each other across all states, presenting a challenge in actively deciding which oracle to use and in which state. We introduce MAPS and MAPS-SE, a class of policy improvement algorithms that perform imitation learning from multiple suboptimal oracles. In particular, MAPS actively selects which of the oracles to imitate and improve their value function estimates, and MAPS-SE additionally leverages an active state exploration criterion to determine which states one should explore. We provide a comprehensive theoretical analysis and demonstrate that MAPS and MAPS-SE enjoy sample efficiency advantage over the state-of-the-art policy improvement algorithms. Empirical results show that MAPS-SE significantly accelerates policy optimization via state-wise imitation learning from multiple oracles across a broad spectrum of control tasks in the DeepMind Control Suite. Our code is publicly available at: https://github.com/ripl/maps.
Authors: Xuefeng Liu, Takuma Yoneda, Rick L. Stevens, Matthew R. Walter, Yuxin Chen
Abstract: While reinforcement learning (RL) has shown promising performance, its sample complexity continues to be a substantial hurdle, restricting its broader application across a variety of domains. Imitation learning (IL) utilizes oracles to improve sample efficiency, yet it is often constrained by the quality of the oracles deployed. which actively interleaves between IL and RL based on an online estimate of their performance. RPI draws on the strengths of IL, using oracle queries to facilitate exploration, an aspect that is notably challenging in sparse-reward RL, particularly during the early stages of learning. As learning unfolds, RPI gradually transitions to RL, effectively treating the learned policy as an improved oracle. This algorithm is capable of learning from and improving upon a diverse set of black-box oracles. Integral to RPI are Robust Active Policy Selection (RAPS) and Robust Policy Gradient (RPG), both of which reason over whether to perform state-wise imitation from the oracles or learn from its own value function when the learner's performance surpasses that of the oracles in a specific state. Empirical evaluations and theoretical analysis validate that RPI excels in comparison to existing state-of-the-art methodologies, demonstrating superior performance across various benchmark domains.
Authors: Prateek Yadav, Leshem Choshen, Colin Raffel, Mohit Bansal
Abstract: Parameter-efficient fine-tuning (PEFT) techniques make it possible to efficiently adapt a language model to create "expert" models that specialize to new tasks or domains. Recent techniques in model merging and compositional generalization leverage these expert models by dynamically composing modules to improve zero/few-shot generalization. Despite the efficiency of PEFT methods, the size of expert models can make it onerous to retrieve expert models per query over high-latency networks like the Internet or serve multiple experts on a single GPU. To address these issues, we present ComPEFT, a novel method for compressing fine-tuning residuals (task vectors) of PEFT based models. ComPEFT employs sparsification and ternary quantization to reduce the size of the PEFT module without performing any additional retraining while preserving or enhancing model performance. In extensive evaluation across T5, T0, and LLaMA-based models with 200M - 65B parameters, ComPEFT achieves compression ratios of 8x - 50x. In particular, we show that ComPEFT improves with scale - stronger models exhibit higher compressibility and better performance. For example, we show that ComPEFT applied to LLaMA outperforms QLoRA by 4.16% on MMLU with a storage size reduction of up to 26x. In addition, we show that the compressed experts produced by ComPEFT maintain few-shot compositional generalization capabilities, facilitate efficient communication and computation, and exhibit enhanced performance when merged. Lastly, we provide an analysis of different method components, compare it with other PEFT methods, and test ComPEFT's efficacy for compressing the residual of full-finetuning. Our code is available at https://github.com/prateeky2806/compeft.
Authors: Zihan Zhang, Wenhao Zhan, Yuxin Chen, Simon S. Du, Jason D. Lee
Abstract: Multi-distribution learning (MDL), which seeks to learn a shared model that minimizes the worst-case risk across $k$ distinct data distributions, has emerged as a unified framework in response to the evolving demand for robustness, fairness, multi-group collaboration, etc. Achieving data-efficient MDL necessitates adaptive sampling, also called on-demand sampling, throughout the learning process. However, there exist substantial gaps between the state-of-the-art upper and lower bounds on the optimal sample complexity. Focusing on a hypothesis class of Vapnik-Chervonenkis (VC) dimension d, we propose a novel algorithm that yields an varepsilon-optimal randomized hypothesis with a sample complexity on the order of (d+k)/varepsilon^2 (modulo some logarithmic factor), matching the best-known lower bound. Our algorithmic ideas and theory are further extended to accommodate Rademacher classes. The proposed algorithms are oracle-efficient, which access the hypothesis class solely through an empirical risk minimization oracle. Additionally, we establish the necessity of randomization, revealing a large sample size barrier when only deterministic hypotheses are permitted. These findings resolve three open problems presented in COLT 2023 (i.e., citet[Problems 1, 3 and 4]{awasthi2023sample}).
Authors: Jian Xu, Delu Zeng
Abstract: The theory of Bayesian learning incorporates the use of Student-t Processes to model heavy-tailed distributions and datasets with outliers. However, despite Student-t Processes having a similar computational complexity as Gaussian Processes, there has been limited emphasis on the sparse representation of this model. This is mainly due to the increased difficulty in modeling and computation compared to previous sparse Gaussian Processes. Our motivation is to address the need for a sparse representation framework that reduces computational complexity, allowing Student-t Processes to be more flexible for real-world datasets. To achieve this, we leverage the conditional distribution of Student-t Processes to introduce sparse inducing points. Bayesian methods and variational inference are then utilized to derive a well-defined lower bound, facilitating more efficient optimization of our model through stochastic gradient descent. We propose two methods for computing the variational lower bound, one utilizing Monte Carlo sampling and the other employing Jensen's inequality to compute the KL regularization term in the loss function. We propose adopting these approaches as viable alternatives to Gaussian processes when the data might contain outliers or exhibit heavy-tailed behavior, and we provide specific recommendations for their applicability. We evaluate the two proposed approaches on various synthetic and real-world datasets from UCI and Kaggle, demonstrating their effectiveness compared to baseline methods in terms of computational complexity and accuracy, as well as their robustness to outliers.
Authors: Martin Rouault, R\'emi Bardenet, Myl\`ene Ma\"ida
Abstract: Kernel herding belongs to a family of deterministic quadratures that seek to minimize the worst-case integration error over a reproducing kernel Hilbert space (RKHS). These quadrature rules come with strong experimental evidence that this worst-case error decreases at a faster rate than the standard square root of the number of quadrature nodes. This conjectured fast rate is key for integrating expensive-to-evaluate functions, as in Bayesian inference of expensive models, and makes up for the increased computational cost of sampling, compared to i.i.d. or MCMC quadratures. However, there is little theoretical support for this faster-than-square-root rate, at least in the usual case where the RKHS is infinite-dimensional, while recent progress on distribution compression suggests that results on the direct minimization of worst-case integration are possible. In this paper, we study a joint probability distribution over quadrature nodes, whose support tends to minimize the same worst-case error as kernel herding. Our main contribution is to prove that it does outperform i.i.d Monte Carlo, in the sense of coming with a tighter concentration inequality on the worst-case integration error. This first step towards proving a fast error decay demonstrates that the mathematical toolbox developed around Gibbs measures can help understand to what extent kernel herding and its variants improve on computationally cheaper methods. Moreover, we investigate the computational bottlenecks of approximately sampling our quadrature, and we demonstrate on toy examples that a faster rate of convergence, though not worst-case, is likely.
Authors: Giuseppe Canonaco, Leo Ardon, Alberto Pozanco, Daniel Borrajo
Abstract: The use of Potential-Based Reward Shaping (PBRS) has shown great promise in the ongoing research effort to tackle sample inefficiency in Reinforcement Learning (RL). However, choosing the right potential function remains an open challenge. Additionally, RL techniques are usually constrained to use a finite horizon for computational limitations, which introduces a bias when using PBRS. In this paper, we first build some theoretically-grounded intuition on why selecting the potential function as the optimal value function of the task at hand produces performance advantages. We then analyse the bias induced by finite horizons in the context of PBRS producing novel insights. Finally, leveraging abstractions as a way to approximate the optimal value function of the given task, we assess the sample efficiency and performance impact of PBRS on four environments including a goal-oriented navigation task and three Arcade Learning Environments (ALE) games. Remarkably, experimental results show that we can reach the same level of performance as CNN-based solutions with a simple fully-connected network.
Authors: Chih-Hong Cheng, Changshun Wu, Xingyu Zhao, Saddek Bensalem, Harald Ruess
Abstract: The deployment of generative AI (GenAI) models raises significant fairness concerns, addressed in this paper through novel characterization and enforcement techniques specific to GenAI. Unlike standard AI performing specific tasks, GenAI's broad functionality requires ``conditional fairness'' tailored to the context being generated, such as demographic fairness in generating images of poor people versus successful business leaders. We define two fairness levels: the first evaluates fairness in generated outputs, independent of prompts and models; the second assesses inherent fairness with neutral prompts. Given the complexity of GenAI and challenges in fairness specifications, we focus on bounding the worst case, considering a GenAI system unfair if the distance between appearances of a specific group exceeds preset thresholds. We also explore combinatorial testing for assessing relative completeness in intersectional fairness. By bounding the worst case, we develop a prompt injection scheme within an agent-based framework to enforce conditional fairness with minimal intervention, validated on state-of-the-art GenAI systems.
Authors: Shay Deutsch, Lionel Yelibi, Alex Tong Lin, Arjun Ravi Kannan
Abstract: Deriving meaningful representations from complex, high-dimensional data in unsupervised settings is crucial across diverse machine learning applications. This paper introduces a framework for multi-scale graph network embedding based on spectral graph wavelets that employs a contrastive learning approach. We theoretically show that in Paley-Wiener spaces on combinatorial graphs, the spectral graph wavelets operator provides greater flexibility and control over smoothness compared to the Laplacian operator, motivating our approach. A key advantage of the proposed embedding is its ability to establish a correspondence between the embedding and input feature spaces, enabling the derivation of feature importance. We validate the effectiveness of our graph embedding framework on multiple public datasets across various downstream tasks, including clustering and unsupervised feature importance.
Authors: Vignesh Kothapalli, Tianyu Pang, Shenyang Deng, Zongmin Liu, Yaoqing Yang
Abstract: Training strategies for modern deep neural networks (NNs) tend to induce a heavy-tailed (HT) empirical spectral density (ESD) in the layer weights. While previous efforts have shown that the HT phenomenon correlates with good generalization in large NNs, a theoretical explanation of its occurrence is still lacking. Especially, understanding the conditions which lead to this phenomenon can shed light on the interplay between generalization and weight spectra. Our work aims to bridge this gap by presenting a simple, rich setting to model the emergence of HT ESD. In particular, we present a theory-informed setup for 'crafting' heavy tails in the ESD of two-layer NNs and present a systematic analysis of the HT ESD emergence without any gradient noise. This is the first work to analyze a noise-free setting, and we also incorporate optimizer (GD/Adam) dependent (large) learning rates into the HT ESD analysis. Our results highlight the role of learning rates on the Bulk+Spike and HT shape of the ESDs in the early phase of training, which can facilitate generalization in the two-layer NN. These observations shed light on the behavior of large-scale NNs, albeit in a much simpler setting.
Authors: Prateek Yadav, Colin Raffel, Mohammed Muqeeth, Lucas Caccia, Haokun Liu, Tianlong Chen, Mohit Bansal, Leshem Choshen, Alessandro Sordoni
Abstract: The availability of performant pre-trained models has led to a proliferation of fine-tuned expert models that are specialized to a particular domain or task. Model MoErging methods aim to recycle expert models to create an aggregate system with improved performance or generalization. A key component of MoErging methods is the creation of a router that decides which expert model(s) to use for a particular input or application. The promise, effectiveness, and large design space of MoErging has spurred the development of many new methods over the past few years. This rapid pace of development has made it challenging to compare different MoErging methods, which are rarely compared to one another and are often validated in different experimental setups. To remedy such gaps, we present a comprehensive survey of MoErging methods that includes a novel taxonomy for cataloging key design choices and clarifying suitable applications for each method. Apart from surveying MoErging research, we inventory software tools and applications that make use of MoErging. We additionally discuss related fields of study such as model merging, multitask learning, and mixture-of-experts models. Taken as a whole, our survey provides a unified overview of existing MoErging methods and creates a solid foundation for future work in this burgeoning field.
Authors: Xuefeng Gao, Jiale Zha, Xun Yu Zhou
Abstract: We propose a new reinforcement learning (RL) formulation for training continuous-time score-based diffusion models for generative AI to generate samples that maximize reward functions while keeping the generated distributions close to the unknown target data distributions. Different from most existing studies, ours does not involve any pretrained model for the unknown score functions of the noise-perturbed data distributions, nor does it attempt to learn the score functions. Instead, we formulate the problem as entropy-regularized continuous-time RL and show that the optimal stochastic policy has a Gaussian distribution with a known covariance matrix. Based on this result, we parameterize the mean of Gaussian policies and develop an actor--critic type (little) q-learning algorithm to solve the RL problem. A key ingredient in our algorithm design is to obtain noisy observations from the unknown score function via a ratio estimator. Our formulation can also be adapted to solve pure score-matching and fine-tuning pretrained models. Numerically, we show the effectiveness of our approach by comparing its performance with two state-of-the-art RL methods that fine-tune pretrained models on several generative tasks including high-dimensional image generations. Finally, we discuss extensions of our RL formulation to probability flow ODE implementation of diffusion models and to conditional diffusion models.
Authors: Yixian Shen, Qi Bi, Jia-Hong Huang, Hongyi Zhu, Andy D. Pimentel, Anuj Pathania
Abstract: We present a new adaptation method MaCP, Minimal yet Mighty adaptive Cosine Projection, that achieves exceptional performance while requiring minimal parameters and memory for fine-tuning large foundation models. Its general idea is to exploit the superior energy compaction and decorrelation properties of cosine projection to improve both model efficiency and accuracy. Specifically, it projects the weight change from the low-rank adaptation into the discrete cosine space. Then, the weight change is partitioned over different levels of the discrete cosine spectrum, and each partition's most critical frequency components are selected. Extensive experiments demonstrate the effectiveness of MaCP across a wide range of single-modality tasks, including natural language understanding, natural language generation, text summarization, as well as multi-modality tasks such as image classification and video understanding. MaCP consistently delivers superior accuracy, significantly reduced computational complexity, and lower memory requirements compared to existing alternatives.
Authors: Haoye Chai, Shiyuan Zhang, Xiaoqian Qi, Baohua Qiu, Yong Li
Abstract: Mobile traffic forecasting allows operators to anticipate network dynamics and performance in advance, offering substantial potential for enhancing service quality and improving user experience. However, existing models are often task-oriented and are trained with tailored data, which limits their effectiveness in diverse mobile network tasks of Base Station (BS) deployment, resource allocation, energy optimization, etc. and hinders generalization across different urban environments. Foundation models have made remarkable strides across various domains of NLP and CV due to their multi-tasking adaption and zero/few-shot learning capabilities. In this paper, we propose an innovative Foundation model for Mo}bile traffic forecasting (FoMo), aiming to handle diverse forecasting tasks of short/long-term predictions and distribution generation across multiple cities to support network planning and optimization. FoMo combines diffusion models and transformers, where various spatio-temporal masks are proposed to enable FoMo to learn intrinsic features of different tasks, and a contrastive learning strategy is developed to capture the correlations between mobile traffic and urban contexts, thereby improving its transfer learning capability. Extensive experiments on 9 real-world datasets demonstrate that FoMo outperforms current models concerning diverse forecasting tasks and zero/few-shot learning, showcasing a strong universality.
Authors: Siva Viknesh, Younes Tatari, Chase Christenson, Amirhossein Arzani
Abstract: Identifying dynamical systems characterized by nonlinear parameters presents significant challenges in deriving mathematical models that enhance understanding of physics. Traditional methods, such as Sparse Identification of Nonlinear Dynamics (SINDy) and symbolic regression, can extract governing equations from observational data; however, they also come with distinct advantages and disadvantages. This paper introduces a novel method within the SINDy framework, termed ADAM-SINDy, which synthesizes the strengths of established approaches by employing the ADAM optimization algorithm. This facilitates the simultaneous optimization of nonlinear parameters and coefficients associated with nonlinear candidate functions, enabling precise parameter estimation without requiring prior knowledge of nonlinear characteristics such as trigonometric frequencies, exponential bandwidths, or polynomial exponents, thereby addressing a key limitation of SINDy. Through an integrated global optimization, ADAM-SINDy dynamically adjusts all unknown variables in response to data, resulting in an adaptive identification procedure that reduces the sensitivity to the library of candidate functions. The performance of the ADAM-SINDy methodology is demonstrated across a spectrum of dynamical systems, including benchmark coupled nonlinear ordinary differential equations such as oscillators, chaotic fluid flows, reaction kinetics, pharmacokinetics, as well as nonlinear partial differential equations (wildfire transport). The results demonstrate significant improvements in identifying parameterized dynamical systems and underscore the importance of concurrently optimizing all parameters, particularly those characterized by nonlinear parameters. These findings highlight the potential of ADAM-SINDy to extend the applicability of the SINDy framework in addressing more complex challenges in dynamical system identification.
Authors: Yonatan Kurniawan (Brigham Young University, Provo, UT, USA), Tracianne B. Neilsen (Brigham Young University, Provo, UT, USA), Benjamin L. Francis (Achilles Heel Technologies, Orem, UT, USA), Alex M. Stankovic (SLAC National Accelerator Laboratory, Menlo Park, CA, USA), Mingjian Wen (University of Electronic Science and Technology of China, Chengdu, China), Ilia Nikiforov (University of Minnesota, Minneapolis, MN, USA), Ellad B. Tadmor (University of Minnesota, Minneapolis, MN, USA), Vasily V. Bulatov (Lawrence Livermore National Laboratory), Vincenzo Lordi (Lawrence Livermore National Laboratory), Mark K. Transtrum (Brigham Young University, Provo, UT, USA, SLAC National Accelerator Laboratory, Menlo Park, CA, USA)
Abstract: The efficacy of mathematical models heavily depends on the quality of the training data, yet collecting sufficient data is often expensive and challenging. Many modeling applications require inferring parameters only as a means to predict other quantities of interest (QoI). Because models often contain many unidentifiable (sloppy) parameters, QoIs often depend on a relatively small number of parameter combinations. Therefore, we introduce an information-matching criterion based on the Fisher Information Matrix to select the most informative training data from a candidate pool. This method ensures that the selected data contain sufficient information to learn only those parameters that are needed to constrain downstream QoIs. It is formulated as a convex optimization problem, making it scalable to large models and datasets. We demonstrate the effectiveness of this approach across various modeling problems in diverse scientific fields, including power systems and underwater acoustics. Finally, we use information-matching as a query function within an Active Learning loop for material science applications. In all these applications, we find that a relatively small set of optimal training data can provide the necessary information for achieving precise predictions. These results are encouraging for diverse future applications, particularly active learning in large machine learning models.
Authors: Jan Boelts, Michael Deistler, Manuel Gloeckler, \'Alvaro Tejero-Cantero, Jan-Matthis Lueckmann, Guy Moss, Peter Steinbach, Thomas Moreau, Fabio Muratore, Julia Linhart, Conor Durkan, Julius Vetter, Benjamin Kurt Miller, Maternus Herold, Abolfazl Ziaeemehr, Matthijs Pals, Theo Gruner, Sebastian Bischoff, Nastya Krouglova, Richard Gao, Janne K. Lappalainen, B\'alint Mucs\'anyi, Felix Pei, Auguste Schulz, Zinovia Stefanidi, Pedro Rodrigues, Cornelius Schr\"oder, Faried Abu Zaid, Jonas Beck, Jaivardhan Kapoor, David S. Greenberg, Pedro J. Gon\c{c}alves, Jakob H. Macke
Abstract: Scientists and engineers use simulators to model empirically observed phenomena. However, tuning the parameters of a simulator to ensure its outputs match observed data presents a significant challenge. Simulation-based inference (SBI) addresses this by enabling Bayesian inference for simulators, identifying parameters that match observed data and align with prior knowledge. Unlike traditional Bayesian inference, SBI only needs access to simulations from the model and does not require evaluations of the likelihood function. In addition, SBI algorithms do not require gradients through the simulator, allow for massive parallelization of simulations, and can perform inference for different observations without further simulations or training, thereby amortizing inference. Over the past years, we have developed, maintained, and extended sbi, a PyTorch-based package that implements Bayesian SBI algorithms based on neural networks. The sbi toolkit implements a wide range of inference methods, neural network architectures, sampling methods, and diagnostic tools. In addition, it provides well-tested default settings, but also offers flexibility to fully customize every step of the simulation-based inference workflow. Taken together, the sbi toolkit enables scientists and engineers to apply state-of-the-art SBI methods to black-box simulators, opening up new possibilities for aligning simulations with empirically observed data.
Authors: Dongjae Jeon, Dueun Kim, Albert No
Abstract: In this paper, we introduce a geometric framework to analyze memorization in diffusion models through the sharpness of the log probability density. We mathematically justify a previously proposed score-difference-based memorization metric by demonstrating its effectiveness in quantifying sharpness. Additionally, we propose a novel memorization metric that captures sharpness at the initial stage of image generation in latent diffusion models, offering early insights into potential memorization. Leveraging this metric, we develop a mitigation strategy that optimizes the initial noise of the generation process using a sharpness-aware regularization term.
Authors: Ahmad Mousavi, Ramin Zandvakili
Abstract: Kernel-free quadratic surface support vector machines have recently gained traction due to their flexibility in modeling nonlinear decision boundaries without relying on kernel functions. However, the introduction of a full quadratic classifier significantly increases the number of model parameters, scaling quadratically with data dimensionality, which often leads to overfitting and makes interpretation difficult. To address these challenges, we propose a sparse variant of the QSVM by enforcing a cardinality constraint on the model parameters. While enhancing generalization and promoting sparsity, leveraging the $\ell_0$-norm inevitably incurs additional computational complexity. To tackle this, we develop a penalty decomposition algorithm capable of producing solutions that provably satisfy the first-order Lu-Zhang optimality conditions. Our approach accommodates both hinge and quadratic loss functions. In both cases, we demonstrate that the subproblems arising within the algorithm either admit closed-form solutions or can be solved efficiently through dual formulations, which contributes to the method's overall effectiveness. We also analyze the convergence behavior of the algorithm under both loss settings. Finally, we validate our approach on several real-world datasets, demonstrating its ability to reduce overfitting while maintaining strong classification performance. The complete implementation and experimental code are publicly available at https://github.com/raminzandvakili/L0-QSVM.
Authors: Pablo Barcel\'o, Alexander Kozachinskiy, Tomasz Steifer
Abstract: The notion of \emph{rank} of a Boolean function has been a cornerstone in PAC learning theory, enabling quasipolynomial-time learning algorithms for polynomial-size decision trees. We present a novel characterization of rank, grounded in the well-known Transformer architecture. We show that the rank of a function $f$ corresponds to the minimum number of \emph{Chain of Thought} (CoT) steps required by a single-layer Transformer with hard attention to compute $f$. Based on this characterization we establish tight bounds on the number of CoT steps required for specific problems, showing that \(\ell\)-fold function composition necessitates exactly \(\ell\) CoT steps. Furthermore, we analyze the problem of identifying the position of the \(k\)-th occurrence of 1 in a Boolean sequence, proving that it requires \(k\) CoT steps. Finally, we introduce the notion of the multi-head rank that captures multi-head single-layer transformers, and perform the analysis of PAC-learnability of the classes of functions with bounded multi-head rank.
Authors: Harshwardhan Praveen, Jacob Brown, Christopher Earls
Abstract: In this work, we present a mesh-independent, data-driven library, chebgreen, to mathematically model one-dimensional systems, possessing an associated control parameter, and whose governing partial differential equation is unknown. The proposed method learns an Empirical Green's Function for the associated, but hidden, boundary value problem, in the form of a Rational Neural Network from which we subsequently construct a bivariate representation in a Chebyshev basis. We uncover the Green's function, at an unseen control parameter value, by interpolating the left and right singular functions within a suitable library, expressed as points on a manifold of Quasimatrices, while the associated singular values are interpolated with Lagrange polynomials.
Authors: Natalie Maus, Kyurae Kim, Yimeng Zeng, Haydn Thomas Jones, Fangping Wan, Marcelo Der Torossian Torres, Cesar de la Fuente-Nunez, Jacob R. Gardner
Abstract: In multi-objective black-box optimization, the goal is typically to find solutions that optimize a set of $T$ black-box objective functions, $f_1$, ..., $f_T$, simultaneously. Traditional approaches often seek a single Pareto-optimal set that balances trade-offs among all objectives. In this work, we consider a problem setting that departs from this paradigm: finding a small set of K < T solutions, that collectively "covers" the T objectives. A set of solutions is defined as "covering" if, for each objective $f_1$, ..., $f_T$, there is at least one good solution. A motivating example for this problem setting occurs in drug design. For example, we may have T pathogens and aim to identify a set of K < T antibiotics such that at least one antibiotic can be used to treat each pathogen. To address this problem, we propose Multi-Objective Coverage Bayesian Optimization (MOCOBO), a principled algorithm designed to efficiently find a covering set. We validate our approach through experiments on challenging high-dimensional tasks, including applications in peptide and molecular design, where MOCOBO is shown to find high-performing covering sets of solutions. The results show that the coverage of the K < T solutions found by MOCOBO matches or nearly matches the coverage of T solutions obtained by optimizing each objective individually. Furthermore, in in vitro experiments, the peptides found by MOCOBO exhibited high potency against drug-resistant pathogens, further demonstrating the potential of MOCOBO for drug discovery. We make code available here: https://github.com/nataliemaus/mocobo.
Authors: Xinyi Wu, Yifei Wang, Stefanie Jegelka, Ali Jadbabaie
Abstract: Recent studies have revealed various manifestations of position bias in transformer architectures, from the "lost-in-the-middle" phenomenon to attention sinks, yet a comprehensive theoretical understanding of how attention masks and positional encodings shape these biases remains elusive. This paper presents a graph-theoretic framework for analyzing position bias in multi-layer attention. Modeling attention masks as directed graphs, we quantify how tokens interact with contextual information based on their sequential positions. We uncover two key insights: First, causal masking inherently biases attention toward earlier positions, as tokens in deeper layers attend to increasingly more contextualized representations of earlier tokens. Second, we characterize the competing effects of the causal mask and relative positional encodings, such as the decay mask and rotary positional encoding (RoPE): while both mechanisms introduce distance-based decay within individual attention maps, their aggregate effect across multiple attention layers$\unicode{x2013}$coupled with the causal mask$\unicode{x2013}$leads to a trade-off between the long-term decay effects and the cumulative importance of early sequence positions. Through controlled numerical experiments, we not only validate our theoretical findings but also reproduce position biases observed in real-world LLMs. Our framework offers a principled foundation for understanding positional biases in transformers, shedding light on the complex interplay of attention mechanism components and guiding more informed architectural design.
Authors: Yiye Chen, Harpreet Sawhney, Nicholas Gyd\'e, Yanan Jian, Jack Saunders, Patricio Vela, Ben Lundell
Abstract: Scene graphs have emerged as a structured and serializable environment representation for grounded spatial reasoning with Large Language Models (LLMs). In this work, we propose SG^2, an iterative Schema-Guided Scene-Graph reasoning framework based on multi-agent LLMs. The agents are grouped into two modules: a (1) Reasoner module for abstract task planning and graph information queries generation, and a (2) Retriever module for extracting corresponding graph information based on code-writing following the queries. Two modules collaborate iteratively, enabling sequential reasoning and adaptive attention to graph information. The scene graph schema, prompted to both modules, serves to not only streamline both reasoning and retrieval process, but also guide the cooperation between two modules. This eliminates the need to prompt LLMs with full graph data, reducing the chance of hallucination due to irrelevant information. Through experiments in multiple simulation environments, we show that our framework surpasses existing LLM-based approaches and baseline single-agent, tool-based Reason-while-Retrieve strategy in numerical Q\&A and planning tasks.
Authors: \'Isak P\'etursson, Mar\'ia \'Oskarsd\'ottir
Abstract: Stochastic partial differential equations (SPDEs) describe the evolution of random processes over space and time, but their solutions are often analytically intractable and computationally expensive to estimate. In this paper, we propose the Learned Expectation Collapser (LEC), a physics-informed neural framework designed to approximate the expected value of linear SPDE solutions without requiring domain discretization. By leveraging randomized sampling of both space-time coordinates and noise realizations during training, LEC trains standard feedforward neural networks to minimize residual loss across multiple stochastic samples. We hypothesize and empirically confirm that this training regime drives the network to converge toward the expected value of the solution of the SPDE. Using the stochastic heat equation as a testbed, we evaluate performance across a diverse set of 144 experimental configurations that span multiple spatial dimensions, noise models, and forcing functions. The results show that the model consistently learns accurate approximations of the expected value of the solution in lower dimensions and a predictable decrease in accuracy with increased spatial dimensions, with improved stability and robustness under increased Monte Carlo sampling. Our findings offer new insight into how neural networks implicitly learn statistical structure from stochastic differential operators and suggest a pathway toward scalable, simulator-free SPDE solvers.
Authors: Huchen Yang, Chuanqi Chen, Jin-Long Wu
Abstract: Digital twins have been actively explored in many engineering applications, such as manufacturing and autonomous systems. However, model discrepancy is ubiquitous in most digital twin models and has significant impacts on the performance of using those models. In recent years, data-driven modeling techniques have been demonstrated promising in characterizing the model discrepancy in existing models, while the training data for the learning of model discrepancy is often obtained in an empirical way and an active approach of gathering informative data can potentially benefit the learning of model discrepancy. On the other hand, Bayesian experimental design (BED) provides a systematic approach to gathering the most informative data, but its performance is often negatively impacted by the model discrepancy. In this work, we build on sequential BED and propose an efficient approach to iteratively learn the model discrepancy based on the data from the BED. The performance of the proposed method is validated by a classical numerical example governed by a convection-diffusion equation, for which full BED is still feasible. The proposed method is then further studied in the same numerical example with a high-dimensional model discrepancy, which serves as a demonstration for the scenarios where full BED is not practical anymore. An ensemble-based approximation of information gain is further utilized to assess the data informativeness and to enhance learning model discrepancy. The results show that the proposed method is efficient and robust to the active learning of high-dimensional model discrepancy, using data suggested by the sequential BED. We also demonstrate that the proposed method is compatible with both classical numerical solvers and modern auto-differentiable solvers.
Authors: Can Chen, Jun-Kun Wang
Abstract: The technique of ``testing by betting" frames nonparametric sequential hypothesis testing as a multiple-round game, where a player bets on future observations that arrive in a streaming fashion, accumulates wealth that quantifies evidence against the null hypothesis, and rejects the null once the wealth exceeds a specified threshold while controlling the false positive error. Designing an online learning algorithm that achieves a small regret in the game can help rapidly accumulate the bettor's wealth, which in turn can shorten the time to reject the null hypothesis under the alternative $H_1$. However, many of the existing works employ the Online Newton Step (ONS) to update within a halved decision space to avoid a gradient explosion issue, which is potentially conservative for rapid wealth accumulation. In this paper, we introduce a novel strategy utilizing interior-point methods in optimization that allows updates across the entire interior of the decision space without the risk of gradient explosion. Our approach not only maintains strong statistical guarantees but also facilitates faster null hypothesis rejection, while being as computationally lightweight as ONS thanks to its closed-form updates.
Authors: Xuefeng Liu, Hung T. C. Le, Siyu Chen, Rick Stevens, Zhuoran Yang, Matthew R. Walter, Yuxin Chen
Abstract: Online reinforcement learning (RL) enhances policies through direct interactions with the environment, but faces challenges related to sample efficiency. In contrast, offline RL leverages extensive pre-collected data to learn policies, but often produces suboptimal results due to limited data coverage. Recent efforts integrate offline and online RL in order to harness the advantages of both approaches. However, effectively combining online and offline RL remains challenging due to issues that include catastrophic forgetting, lack of robustness to data quality and limited sample efficiency in data utilization. In an effort to address these challenges, we introduce A3RL, which incorporates a novel confidence aware Active Advantage Aligned (A3) sampling strategy that dynamically prioritizes data aligned with the policy's evolving needs from both online and offline sources, optimizing policy improvement. Moreover, we provide theoretical insights into the effectiveness of our active sampling strategy and conduct diverse empirical experiments and ablation studies, demonstrating that our method outperforms competing online RL techniques that leverage offline data.
Authors: Sophia Sklaviadis, Andre Martins, Mario Figueiredo
Abstract: From a variational perspective, many statistical learning criteria involve seeking a distribution that balances empirical risk and regularization. In this paper, we broaden this perspective by introducing a new general class of variational methods based on Fenchel-Young (FY) losses, treated as divergences that generalize (and encompass) the familiar Kullback-Leibler divergence at the core of classical variational learning. Our proposed formulation -- FY variational learning -- includes as key ingredients new notions of FY free energy, FY evidence, FY evidence lower bound, and FY posterior. We derive alternating minimization and gradient backpropagation algorithms to compute (or lower bound) the FY evidence, which enables learning a wider class of models than previous variational formulations. This leads to generalized FY variants of classical algorithms, such as an FY expectation-maximization (FYEM) algorithm, and latent-variable models, such as an FY variational autoencoder (FYVAE). Our new methods are shown to be empirically competitive, often outperforming their classical counterparts, and most importantly, to have qualitatively novel features. For example, FYEM has an adaptively sparse E-step, while the FYVAE can support models with sparse observations and sparse posteriors.
Authors: Jaike van Twiller, Yossiri Adulyasak, Erick Delage, Djordje Grbic, Rune M{\o}ller Jensen
Abstract: Reinforcement learning (RL) has shown promise in solving various combinatorial optimization problems. However, conventional RL faces challenges when dealing with complex, real-world constraints, especially when action space feasibility is explicit and dependent on the corresponding state or trajectory. In this work, we address stochastic sequential dynamic decision-making problems with state-dependent constraints. As a relevant and real-world case study, we focus on the master stowage planning problem in container shipping, which aims to optimize revenue and operational costs under demand uncertainty and operational constraints. We propose a deep RL framework with an encoder-decoder model and feasibility layers that satisfy convex constraints and maintain unbiased gradient flow, which embed problem instances, current solutions, and demand uncertainty to guide learning. Experiments show that our model efficiently finds adaptive, feasible solutions that generalize across varying distributions and scale to larger instances, outperforming state-of-the-art baselines in constrained RL and stochastic programming. By uniting artificial intelligence and operations research, our policy empowers humans to make adaptive, uncertainty-aware decisions for resilient and sustainable planning.
Authors: Lucas Torroba-Hennigen, Hunter Lang, Han Guo, Yoon Kim
Abstract: We study memory-efficient optimization of neural networks (in particular language models) with linear gradient transformations, where the gradients are linearly mapped to a lower dimensional space than the full parameter space, thus saving memory required for gradient accumulation and optimizer state persistence. The model parameters are updated by first performing an optimization step in the lower dimensional space and then going back into the original parameter space via the linear map's transpose. We show that optimizing the model in this transformed space is equivalent to reparameterizing the original model through a linear adapter that additively modifies the model parameters, and then only optimizing the adapter's parameters. When the transformation is Kronecker-factored, this establishes an equivalence between GaLore and one-sided LoRA. We show that this duality between gradient transformations and adapter-based reparameterizations unifies existing approaches to memory-efficient training and suggests new techniques for improving training efficiency and memory use.
Authors: Vlad Sobal, Wancong Zhang, Kyunghyun Cho, Randall Balestriero, Tim G. J. Rudner, Yann LeCun
Abstract: A long-standing goal in AI is to build agents that can solve a variety of tasks across different environments, including previously unseen ones. Two dominant approaches tackle this challenge: (i) reinforcement learning (RL), which learns policies through trial and error, and (ii) optimal control, which plans actions using a learned or known dynamics model. However, their relative strengths and weaknesses remain underexplored in the setting where agents must learn from offline trajectories without reward annotations. In this work, we systematically analyze the performance of different RL and control-based methods under datasets of varying quality. On the RL side, we consider goal-conditioned and zero-shot approaches. On the control side, we train a latent dynamics model using the Joint Embedding Predictive Architecture (JEPA) and use it for planning. We study how dataset properties-such as data diversity, trajectory quality, and environment variability-affect the performance of these approaches. Our results show that model-free RL excels when abundant, high-quality data is available, while model-based planning excels in generalization to novel environment layouts, trajectory stitching, and data-efficiency. Notably, planning with a latent dynamics model emerges as a promising approach for zero-shot generalization from suboptimal data.
Authors: Amartaivan Sanjjamts, Hiroshi Morita, Togootogtokh Enkhtogtokh
Abstract: Understanding collective pedestrian movement is crucial for applications in crowd management, autonomous navigation, and human-robot interaction. This paper investigates the use of sequential deep learning models, including Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) networks, and Transformers, for real-time flock detection in multi-pedestrian trajectories. Our proposed approach consists of a two-stage process: first, a pre-trained binary classification model is used for pairwise trajectory classification, and second, the learned representations are applied to identify multi-agent flocks dynamically. We validate our method using real-world group movement datasets, demonstrating its robustness across varying sequence lengths and diverse movement patterns. Experimental results indicate that our model consistently detects pedestrian flocks with high accuracy and stability, even in dynamic and noisy environments. Furthermore, we extend our approach to identify other forms of collective motion, such as convoys and swarms, paving the way for more comprehensive multi-agent behavior analysis.
Authors: Ruth Crasto
Abstract: Geographic distribution shift arises when the distribution of locations on Earth in a training dataset is different from what is seen at test time. The most common approaches to tackling geographic distribution shift treat regions delimited by administrative boundaries such as countries or continents as separate domains and apply standard domain adaptation methods, ignoring geographic coordinates that are often available as metadata. This paper proposes the use of location encoders for modeling continuous, learnable domain assignment. We show how both non-parametric sine-cosine encoders and pre-trained location encoders can be used in conjunction with standard domain adaptation methods for improved robustness to geographic distribution shift. Our proposed methods achieve new state-of-the-art results on two geo-tagged remote sensing datasets from the WILDS benchmark. We have made our code publicly available at: https://github.com/crastoru/wilds-geoshift.
Authors: Fengdi Che, Bryan Chan, Chen Ma, A. Rupam Mahmood
Abstract: Off-policy policy evaluation (OPE), an essential component of reinforcement learning, has long suffered from stationary state distribution mismatch, undermining both stability and accuracy of OPE estimates. While existing methods correct distribution shifts by estimating density ratios, they often rely on expensive optimization or backward Bellman-based updates and struggle to outperform simpler baselines. We introduce AVG-DICE, a computationally simple Monte Carlo estimator for the density ratio that averages discounted importance sampling ratios, providing an unbiased and consistent correction. AVG-DICE extends naturally to nonlinear function approximation using regression, which we roughly tune and test on OPE tasks based on Mujoco Gym environments and compare with state-of-the-art density-ratio estimators using their reported hyperparameters. In our experiments, AVG-DICE is at least as accurate as state-of-the-art estimators and sometimes offers orders-of-magnitude improvements. However, a sensitivity analysis shows that best-performing hyperparameters may vary substantially across different discount factors, so a re-tuning is suggested.
Authors: Greg d'Eon, Hala Murad, Kevin Leyton-Brown, James R. Wright
Abstract: Behavioral game theory models serve two purposes: yielding insights into how human decision-making works, and predicting how people would behave in novel strategic settings. A system called GameNet represents the state of the art for predicting human behavior in the setting of unrepeated simultaneous-move games, combining a simple "level-k" model of strategic reasoning with a complex neural network model of non-strategic "level-0" behavior. Although this reliance on well-established ideas from cognitive science ought to make GameNet interpretable, the flexibility of its level-0 model raises the possibility that it is able to emulate strategic reasoning. In this work, we prove that GameNet's level-0 model is indeed too general. We then introduce ElementaryNet, a novel neural network that is provably incapable of expressing strategic behavior. We show that these additional restrictions are empirically harmless, leading ElementaryNet to statistically indistinguishable predictive performance vs GameNet. We then show how it is possible to derive insights about human behavior by varying ElementaryNet's features and interpreting its parameters, finding evidence of iterative reasoning, learning about the depth of this reasoning process, and showing the value of a rich level-0 specification.
Authors: Ihab Asaad, Maha Shadaydeh, Joachim Denzler
Abstract: Machine learning classification models trained with empirical risk minimization (ERM) often inadvertently rely on spurious correlations. When absent in the test data, these unintended associations between non-target attributes and target labels lead to poor generalization. This paper addresses this problem from a model optimization perspective and proposes a novel method, Gradient Extrapolation for Debiased Representation Learning (GERNE), designed to learn debiased representations in both known and unknown attribute training cases. GERNE uses two distinct batches with different amounts of spurious correlations and defines the target gradient as a linear extrapolation of the gradients computed from each batch's loss. Our analysis shows that when the extrapolated gradient points toward the batch gradient with fewer spurious correlations, it effectively guides training toward learning a debiased model. GERNE serves as a general framework for debiasing, encompassing ERM and Resampling methods as special cases. We derive the theoretical upper and lower bounds of the extrapolation factor employed by GERNE. By tuning this factor, GERNE can adapt to maximize either Group-Balanced Accuracy (GBA) or Worst-Group Accuracy (WGA). We validate GERNE on five vision and one NLP benchmarks, demonstrating competitive and often superior performance compared to state-of-the-art baselines. The project page is available at: https://gerne-debias.github.io/.
Authors: Dawood Wasif, Dian Chen, Sindhuja Madabushi, Nithin Alluru, Terrence J. Moore, Jin-Hee Cho
Abstract: Federated Learning (FL) enables collaborative model training while preserving data privacy; however, balancing privacy preservation (PP) and fairness poses significant challenges. In this paper, we present the first unified large-scale empirical study of privacy-fairness-utility trade-offs in FL, advancing toward responsible AI deployment. Specifically, we systematically compare Differential Privacy (DP), Homomorphic Encryption (HE), and Secure Multi-Party Computation (SMC) with fairness-aware optimizers including q-FedAvg, q-MAML, Ditto, evaluating their performance under IID and non-IID scenarios using benchmark (MNIST, Fashion-MNIST) and real-world datasets (Alzheimer's MRI, credit-card fraud detection). Our analysis reveals HE and SMC significantly outperform DP in achieving equitable outcomes under data skew, although at higher computational costs. Remarkably, we uncover unexpected interactions: DP mechanisms can negatively impact fairness, and fairness-aware optimizers can inadvertently reduce privacy effectiveness. We conclude with practical guidelines for designing robust FL systems that deliver equitable, privacy-preserving, and accurate outcomes.
Authors: Jeremy Diamzon, Daniele Venturi
Abstract: We develop new uncertainty propagation methods for feed-forward neural network architectures with leaky ReLU activation functions subject to random perturbations in the input vectors. In particular, we derive analytical expressions for the probability density function (PDF) of the neural network output and its statistical moments as a function of the input uncertainty and the parameters of the network, i.e., weights and biases. A key finding is that an appropriate linearization of the leaky ReLU activation function yields accurate statistical results even for large perturbations in the input vectors. This can be attributed to the way information propagates through the network. We also propose new analytically tractable Gaussian copula surrogate models to approximate the full joint PDF of the neural network output. To validate our theoretical results, we conduct Monte Carlo simulations and a thorough error analysis on a multi-layer neural network representing a nonlinear integro-differential operator between two polynomial function spaces. Our findings demonstrate excellent agreement between the theoretical predictions and Monte Carlo simulations.
Authors: Xuefeng Liu, Songhao Jiang, Chih-chan Tien, Jinbo Xu, Rick Stevens
Abstract: Protein representation learning is critical for numerous biological tasks. Recently, large transformer-based protein language models (pLMs) pretrained on large scale protein sequences have demonstrated significant success in sequence-based tasks. However, pLMs lack structural context. Conversely, graph neural networks (GNNs) designed to leverage 3D structural information have shown promising generalization in protein-related prediction tasks, but their effectiveness is often constrained by the scarcity of labeled structural data. Recognizing that sequence and structural representations are complementary perspectives of the same protein entity, we propose a multimodal bidirectional hierarchical fusion framework to effectively merge these modalities. Our framework employs attention and gating mechanisms to enable effective interaction between pLMs-generated sequential representations and GNN-extracted structural features, improving information exchange and enhancement across layers of the neural network. This bidirectional and hierarchical (Bi-Hierarchical) fusion approach leverages the strengths of both modalities to capture richer and more comprehensive protein representations. Based on the framework, we further introduce local Bi-Hierarchical Fusion with gating and global Bi-Hierarchical Fusion with multihead self-attention approaches. Our method demonstrates consistent improvements over strong baselines and existing fusion techniques in a variety of protein representation learning benchmarks, including enzyme EC classification, model quality assessment, protein-ligand binding affinity prediction, protein-protein binding site prediction, and B cell epitopes prediction. Our method establishes a new state-of-the-art for multimodal protein representation learning, emphasizing the efficacy of Bi-Hierarchical Fusion in bridging sequence and structural modalities.
Authors: Zhang Xi-Jia, Yue Guo, Shufei Chen, Simon Stepputtis, Matthew Gombolay, Katia Sycara, Joseph Campbell
Abstract: Intelligent agents, such as robots, are increasingly deployed in real-world, human-centric environments. To foster appropriate human trust and meet legal and ethical standards, these agents must be able to explain their behavior. However, state-of-the-art agents are typically driven by black-box models like deep neural networks, limiting their interpretability. We propose a method for generating natural language explanations of agent behavior based only on observed states and actions -- without access to the agent's underlying model. Our approach learns a locally interpretable surrogate model of the agent's behavior from observations, which then guides a large language model to generate plausible explanations with minimal hallucination. Empirical results show that our method produces explanations that are more comprehensible and correct than those from baselines, as judged by both language models and human evaluators. Furthermore, we find that participants in a user study more accurately predicted the agent's future actions when given our explanations, suggesting improved understanding of agent behavior.
Authors: Lunyiu Nie, Zhimin Ding, Kevin Yu, Marco Cheung, Chris Jermaine, Swarat Chaudhuri
Abstract: The inference-time resource costs of large language and vision models present a growing challenge in production deployments. We propose the use of foundation model programs, i.e., programs that can invoke foundation models with varying resource costs and performance, as an approach to this problem. Specifically, we present a method that translates a task into a program, then learns a policy for resource allocation that, on each input, selects foundation model "backends" for each program module. The policy uses smaller, cheaper backends to handle simpler subtasks, while allowing more complex subtasks to leverage larger, more capable models. We evaluate the method on two new "streaming" visual question-answering tasks in which a system answers a question on a sequence of inputs, receiving ground-truth feedback after each answer. Compared to monolithic multi-modal models, our implementation achieves up to 98% resource savings with minimal accuracy loss, demonstrating its potential for scalable and resource-efficient multi-modal inference.
Authors: Matthew B. Webster, Dongheon Lee, Joonnyong Lee
Abstract: Biosignals can be viewed as mixtures measuring particular physiological events, and blind source separation (BSS) aims to extract underlying source signals from mixtures. This paper proposes a self-supervised multi-encoder autoencoder (MEAE) to separate heartbeat-related source signals from photoplethysmogram (PPG), enhancing heart rate (HR) detection in noisy PPG data. The MEAE is trained on PPG signals from a large open polysomnography database without any pre-processing or data selection. The trained network is then applied to a noisy PPG dataset collected during the daily activities of nine subjects. The extracted heartbeat-related source signal significantly improves HR detection as compared to the original PPG. The absence of pre-processing and the self-supervised nature of the proposed method, combined with its strong performance, highlight the potential of MEAE for BSS in biosignal analysis.
Authors: Wei Wang, Maryam Hakimzadeh, Haihui Ruan, Somdatta Goswami
Abstract: Numerical solvers for PDEs often struggle to balance computational cost with accuracy, especially in multiscale and time-dependent systems. Neural operators offer a promising way to accelerate simulations, but their practical deployment is hindered by several challenges: they typically require large volumes of training data generated from high-fidelity solvers, tend to accumulate errors over time in dynamical settings, and often exhibit poor generalization in multiphysics scenarios. This work introduces a novel hybrid framework that integrates physics-informed deep operator network with FEM through domain decomposition and leverages numerical analysis for time marching. Our innovation lies in efficient coupling FE and DeepONet subdomains via a Schwarz method, expecting to solve complex and nonlinear regions by a pretrained DeepONet, while the remainder is handled by conventional FE. To address the challenges of dynamic systems, we embed a time stepping scheme directly into the DeepONet, substantially reducing long-term error propagation. Furthermore, an adaptive subdomain evolution strategy enables the ML-resolved region to expand dynamically, capturing fine-scale features without remeshing. Our framework shows accelerated convergence rates (up to 20% improvement in convergence rates compared to conventional FE coupling approaches) while preserving solution fidelity with error margins consistently below 3%. Our study shows that our proposed hybrid solver: (1) reduces computational costs by eliminating fine mesh requirements, (2) mitigates error accumulation in time-dependent simulations, and (3) enables automatic adaptation to evolving physical phenomena. This work establishes a new paradigm for coupling state of the art physics based and machine learning solvers in a unified framework, offering a robust, reliable, and scalable pathway for high fidelity multiscale simulations.
Authors: Raghavv Goel, Junyoung Park, Mukul Gagrani, Dalton Jones, Matthew Morse, Harper Langston, Mingu Lee, Chris Lott
Abstract: While long context support of large language models has extended their abilities, it also incurs challenges in memory and compute which becomes crucial bottlenecks in resource-restricted devices. Token eviction, a widely adopted post-training methodology designed to alleviate the bottlenecks by evicting less important tokens from the cache, typically uses attention scores as proxy metrics for token importance. However, one major limitation of attention score as a token-wise importance metrics is that it lacks the information about contribution of tokens to the attention output. In this paper, we propose a simple eviction criterion based on the contribution of cached tokens to attention outputs. Our method, CAOTE, optimizes for eviction error due to token eviction, by seamlessly integrating attention scores and value vectors. This is the first method which uses value vector information on top of attention-based eviction scores. Additionally, CAOTE can act as a meta-heuristic method with flexible usage with any token eviction method. We show that CAOTE, when combined with the state-of-the-art attention score-based methods, always improves accuracies on the downstream task, indicating the importance of leveraging information from values during token eviction process.
Authors: Yvonne Jenniges, Maike Sonnewald, Sebastian Maneth, Are Olsen, Boris P. Koch
Abstract: Defining ocean regions and water masses helps to understand marine processes and can serve downstream tasks such as defining marine protected areas. However, such definitions often result from subjective decisions potentially producing misleading, unreproducible outcomes. Here, the aim was to objectively define regions of the North Atlantic through systematic comparison of clustering methods within the Native Emergent Manifold Interrogation (NEMI) framework (Sonnewald, 2023). About 300 million measured salinity, temperature, and oxygen, nitrate, phosphate and silicate concentration values served as input for various clustering methods (k-Means, agglomerative Ward, and Density-Based Spatial Clustering of Applications with Noise (DBSCAN)). Uniform Manifold Approximation and Projection (UMAP) emphasised (dis-)similarities in the data while reducing dimensionality. Based on systematic validation of clustering methods and their hyperparameters using internal, external and relative validation techniques, results showed that UMAP-DBSCAN best represented the data. Strikingly, internal validation metrics proved systematically unreliable for comparing clustering methods. To address stochastic variability, 100 UMAP-DBSCAN clustering runs were conducted and aggregated following NEMI, yielding a final set of 321 clusters. Reproducibility was evaluated via ensemble overlap ($88.81\pm1.8\%$) and mean grid cell-wise uncertainty ($15.49\pm20\%$). Case studies of the Mediterranean Sea, deep Atlantic waters and Labrador Sea showed strong agreement with common water mass definitions. This study revealed a more detailed regionalisation compared to previous concepts such as the Longhurst provinces through systematic clustering method comparison. The applied method is objective, efficient and reproducible and will support future research on biogeochemical differences and changes in oceanic regions.
Authors: Rong Cheng, Jinyi Liu, Yan Zheng, Fei Ni, Jiazhen Du, Hangyu Mao, Fuzheng Zhang, Bo Wang, Jianye Hao
Abstract: Multi-Hop Question Answering (MHQA) tasks permeate real-world applications, posing challenges in orchestrating multi-step reasoning across diverse knowledge domains. While existing approaches have been improved with iterative retrieval, they still struggle to identify and organize dynamic knowledge. To address this, we propose DualRAG, a synergistic dual-process framework that seamlessly integrates reasoning and retrieval. DualRAG operates through two tightly coupled processes: Reasoning-augmented Querying (RaQ) and progressive Knowledge Aggregation (pKA). They work in concert: as RaQ navigates the reasoning path and generates targeted queries, pKA ensures that newly acquired knowledge is systematically integrated to support coherent reasoning. This creates a virtuous cycle of knowledge enrichment and reasoning refinement. Through targeted fine-tuning, DualRAG preserves its sophisticated reasoning and retrieval capabilities even in smaller-scale models, demonstrating its versatility and core advantages across different scales. Extensive experiments demonstrate that this dual-process approach substantially improves answer accuracy and coherence, approaching, and in some cases surpassing, the performance achieved with oracle knowledge access. These results establish DualRAG as a robust and efficient solution for complex multi-hop reasoning tasks.
Authors: Zhaoyan Wang, Xiangchi Song, In-Young Ko
Abstract: Notably, current intelligent transportation systems rely heavily on accurate traffic forecasting and swift inference provision to make timely decisions. While Graph Convolutional Networks (GCNs) have shown benefits in modeling complex traffic dependencies, the existing GCN-based approaches cannot fully extract and fuse multi-granular spatiotemporal features across various spatial and temporal scales sufficiently in a complete manner, proven to yield less accurate results. Besides, as extracting multi-granular features across scales has been a promising strategy across domains such as computer vision, natural language processing, and time-series forecasting, pioneering studies have attempted to leverage a similar mechanism for spatiotemporal traffic data mining. However, additional feature extraction branches introduced in prior studies critically increased model complexity and extended inference time, making it challenging to provide fast forecasts. In this paper, we propose MultiGran-STGCNFog, an efficient fog distributed inference system with a novel traffic forecasting model that employs multi-granular spatiotemporal feature fusion on generated dynamic traffic graphs to fully capture interdependent traffic dynamics. The proposed scheduling algorithm GA-DPHDS, optimizing layer execution order and layer-device scheduling scheme simultaneously, contributes to considerable inference throughput improvement by coordinating heterogeneous fog devices in a pipelined manner. Extensive experiments on real-world datasets demonstrate the superiority of the proposed method over selected GCN baselines.
Authors: Hongze Li, Zesheng Zhou, Zhenbiao Cao, Xinhui Li, Wei Chen, Xiaojin Zhang
Abstract: Traditional Federated Domain Generalization (FedDG) methods focus on learning domain-invariant features or adapting to unseen target domains, often overlooking the unique knowledge embedded within the source domain, especially in strictly isolated federated learning environments. Through experimentation, we discovered a counterintuitive phenomenon.: features learned from a complete source domain have superior generalization capabilities compared to those learned directly from the target domain. This insight leads us to propose the Federated Source Domain Awareness Framework (FedSDAF), the first systematic approach to enhance FedDG by leveraging source domain-aware features. FedSDAF employs a dual-adapter architecture that decouples "local expertise" from "global generalization consensus". A Domain-Aware Adapter, retained locally, extracts and protects the unique discriminative knowledge of each source domain, while a Domain-Invariant Adapter, shared across clients, builds a robust global consensus. To enable knowledge exchange, we introduce a Bidirectional Knowledge Distillation mechanism that facilitates efficient dialogue between the adapters. Extensive experiments on four benchmark datasets (OfficeHome, PACS, VLCS, DomainNet) show that FedSDAF significantly outperforms existing FedDG methods.The source code is available at https://github.com/pizzareapers/FedSDAF.
Authors: Eleftherios Tzanis, Michail E. Klontzas
Abstract: Agentic systems built on large language models (LLMs) offer promising capabilities for automating complex workflows in healthcare AI. We introduce mAIstro, an open-source, autonomous multi-agentic framework for end-to-end development and deployment of medical AI models. The system orchestrates exploratory data analysis, radiomic feature extraction, image segmentation, classification, and regression through a natural language interface, requiring no coding from the user. Built on a modular architecture, mAIstro supports both open- and closed-source LLMs, and was evaluated using a large and diverse set of prompts across 16 open-source datasets, covering a wide range of imaging modalities, anatomical regions, and data types. The agents successfully executed all tasks, producing interpretable outputs and validated models. This work presents the first agentic framework capable of unifying data analysis, AI model development, and inference across varied healthcare applications, offering a reproducible and extensible foundation for clinical and research AI integration. The code is available at: https://github.com/eltzanis/mAIstro
Authors: Gabriel S. Gama, Valdir Grassi Jr
Abstract: Specialized Multi-Task Optimizers (SMTOs) balance task learning in Multi-Task Learning by addressing issues like conflicting gradients and differing gradient norms, which hinder equal-weighted task training. However, recent critiques suggest that equally weighted tasks can achieve competitive results compared to SMTOs, arguing that previous SMTO results were influenced by poor hyperparameter optimization and lack of regularization. In this work, we evaluate these claims through an extensive empirical evaluation of SMTOs, including some of the latest methods, on more complex multi-task problems to clarify this behavior. Our findings indicate that SMTOs perform well compared to uniform loss and that fixed weights can achieve competitive performance compared to SMTOs. Furthermore, we demonstrate why uniform loss perform similarly to SMTOs in some instances. The source code is available at https://github.com/Gabriel-SGama/UnitScal_vs_SMTOs.
Authors: Xvyuan Liu, Xiangfei Qiu, Xingjian Wu, Zhengyu Li, Chenjuan Guo, Jilin Hu, Bin Yang
Abstract: The forecasting of irregular multivariate time series (IMTS) is a critical task in domains like healthcare and climate science. However, this task faces two significant hurdles: 1) the inherent non-uniformity and missing data in IMTS complicate the modeling of temporal dynamics, and 2) existing methods often rely on computationally expensive architectures. To address these dual challenges, we introduce APN, a general and efficient forecasting framework. At the core of APN is a novel Time-Aware Patch Aggregation (TAPA) module that introduces an aggregation-based paradigm for adaptive patching, moving beyond the limitations of fixed-span segmentation and interpolation-based methods. TAPA first learns dynamic temporal boundaries to define data-driven segments. Crucially, instead of resampling or interpolating, it directly computes patch representations via a time-aware weighted aggregation of all raw observations, where weights are determined by each observation's temporal relevance to the segment. This approach provides two key advantages: it preserves data fidelity by avoiding the introduction of artificial data points and ensures complete information coverage by design.The resulting regularized and information-rich patch representations enable the use of a lightweight query module for historical context aggregation and a simple MLP for final prediction. Extensive experiments on multiple real-world datasets demonstrate that APN establishes a new state-of-the-art, significantly outperforming existing methods in both prediction accuracy and computational efficiency.
Authors: Yaxin Hou, Yuheng Jia
Abstract: This paper studies the long-tailed semi-supervised learning (LTSSL) with distribution mismatch, where the class distribution of the labeled training data follows a long-tailed distribution and mismatches with that of the unlabeled training data. Most existing methods introduce auxiliary classifiers (experts) to model various unlabeled data distributions and produce pseudo-labels, but the expertises of various experts are not fully utilized. We observe that different experts are good at predicting different intervals of samples, e.g., long-tailed expert is skilled in samples located in the head interval and uniform expert excels in samples located in the medium interval. Therefore, we propose a dynamic expert assignment module that can estimate the class membership (i.e., head, medium, or tail class) of samples, and dynamically assigns suitable expert to each sample based on the estimated membership to produce high-quality pseudo-label in the training phase and produce prediction in the testing phase. We also theoretically reveal that integrating different experts' strengths will lead to a smaller generalization error bound. Moreover, we find that the deeper features are more biased toward the head class but with more discriminative ability, while the shallower features are less biased but also with less discriminative ability. We, therefore, propose a multi-depth feature fusion module to utilize different depth features to mitigate the model bias. Our method demonstrates its effectiveness through comprehensive experiments on the CIFAR-10-LT, STL-10-LT, and SVHN-LT datasets across various settings.
Authors: Brian Chmiel, Maxim Fishman, Ron Banner, Daniel Soudry
Abstract: We demonstrate, for the first time, fully quantized training (FQT) of large language models (LLMs) using predominantly 4-bit floating-point (FP4) precision for weights, activations, and gradients on datasets up to 200 billion tokens. We extensively investigate key design choices for FP4, including block sizes, scaling formats, and rounding methods. Our analysis shows that the NVFP4 format, where each block of 16 FP4 values (E2M1) shares a scale represented in E4M3, provides optimal results. We use stochastic rounding for backward and update passes and round-to-nearest for the forward pass to enhance stability. Additionally, we identify a theoretical and empirical threshold for effective quantized training: when the gradient norm falls below approximately $\sqrt{3}$ times the quantization noise, quantized training becomes less effective. Leveraging these insights, we successfully train a 7-billion-parameter model on 256 Intel Gaudi2 accelerators. The resulting FP4-trained model achieves downstream task performance comparable to a standard BF16 baseline, confirming that FP4 training is a practical and highly efficient approach for large-scale LLM training. A reference implementation is supplied in https://github.com/Anonymous1252022/fp4-all-the-way .
Authors: Xuandong Zhao, Zhewei Kang, Aosong Feng, Sergey Levine, Dawn Song
Abstract: Training large language models (LLMs) for complex reasoning via Reinforcement Learning with Verifiable Rewards (RLVR) is effective but limited by reliance on costly, domain-specific supervision. We explore Reinforcement Learning from Internal Feedback (RLIF), a framework that enables LLMs to learn from intrinsic signals without external rewards or labeled data. We propose Intuitor, an RLIF method that uses a model's own confidence, termed self-certainty, as its sole reward signal. Intuitor replaces external rewards in Group Relative Policy Optimization (GRPO) with self-certainty scores, enabling fully unsupervised learning. Experiments demonstrate that Intuitor matches GRPO's performance on mathematical benchmarks while achieving superior generalization to out-of-domain tasks like code generation, without requiring gold solutions or test cases. Our findings show that intrinsic model signals can drive effective learning across domains, offering a scalable alternative to RLVR for autonomous AI systems where verifiable rewards are unavailable. Code is available at https://github.com/sunblaze-ucb/Intuitor
Authors: Tao Wu, Jingyuan Chen, Wang Lin, Mengze Li, Yumeng Zhu, Ang Li, Kun Kuang, Fei Wu
Abstract: Large language models (LLMs) are revolutionizing education, with LLM-based agents playing a key role in simulating student behavior. A major challenge in student simulation is modeling the diverse learning patterns of students at various cognitive levels. However, current LLMs, typically trained as ``helpful assistants'', target at generating perfect responses. As a result, they struggle to simulate students with diverse cognitive abilities, as they often produce overly advanced answers, missing the natural imperfections that characterize student learning and resulting in unrealistic simulations. To address this issue, we propose a training-free framework for student simulation. We begin by constructing a cognitive prototype for each student using a knowledge graph, which captures their understanding of concepts from past learning records. This prototype is then mapped to new tasks to predict student performance. Next, we simulate student solutions based on these predictions and iteratively refine them using a beam search method to better replicate realistic mistakes. To validate our approach, we construct the \texttt{Student\_100} dataset, consisting of $100$ students working on Python programming and $5,000$ learning records. Experimental results show that our method consistently outperforms baseline models, achieving $100\%$ improvement in simulation accuracy.
Authors: Yixian Shen, Qi Bi, Jia-Hong Huang, Hongyi Zhu, Andy D. Pimentel, Anuj Pathania
Abstract: We present a new adaptation method MaCP, Minimal yet Mighty adaptive Cosine Projection, that achieves exceptional performance while requiring minimal parameters and memory for fine-tuning large foundation models. Its general idea is to exploit the superior energy compaction and decorrelation properties of cosine projection to improve both model efficiency and accuracy. Specifically, it projects the weight change from the low-rank adaptation into the discrete cosine space. Then, the weight change is partitioned over different levels of the discrete cosine spectrum, and each partition's most critical frequency components are selected. Extensive experiments demonstrate the effectiveness of MaCP across a wide range of single-modality tasks, including natural language understanding, natural language generation, text summarization, as well as multi-modality tasks such as image classification and video understanding. MaCP consistently delivers superior accuracy, significantly reduced computational complexity, and lower memory requirements compared to existing alternatives.
Authors: Adrien Cort\'es, R\'emi Rehm, Victor Letzelter
Abstract: We introduce TimeMCL, a method leveraging the Multiple Choice Learning (MCL) paradigm to forecast multiple plausible time series futures. Our approach employs a neural network with multiple heads and utilizes the Winner-Takes-All (WTA) loss to promote diversity among predictions. MCL has recently gained attention due to its simplicity and ability to address ill-posed and ambiguous tasks. We propose an adaptation of this framework for time-series forecasting, presenting it as an efficient method to predict diverse futures, which we relate to its implicit quantization objective. We provide insights into our approach using synthetic data and evaluate it on real-world time series, demonstrating its promising performance at a light computational cost.
Authors: Zahra Shahrooei, Ali Baheri
Abstract: Deep off-policy actor-critic algorithms have emerged as the leading framework for reinforcement learning in continuous control domains. However, most of these algorithms suffer from poor sample efficiency, especially in environments with sparse rewards. In this paper, we take a step towards addressing this issue by providing a principled directed exploration strategy. We propose Wasserstein Barycenter Soft Actor-Critic (WBSAC) algorithm, which benefits from a pessimistic actor for temporal difference learning and an optimistic actor to promote exploration. This is achieved by using the Wasserstein barycenter of the pessimistic and optimistic policies as the exploration policy and adjusting the degree of exploration throughout the learning process. We compare WBSAC with state-of-the-art off-policy actor-critic algorithms and show that WBSAC is more sample-efficient on MuJoCo continuous control tasks.
Authors: Yichen Luo, Jia Wang, Dapeng Lan, Yu Liu, Zhibo Pang
Abstract: Partial Differential Equations (PDEs) are fundamental for modeling physical systems, yet solving them in a generic and efficient manner using machine learning-based approaches remains challenging due to limited multi-input and multi-scale generalization capabilities, as well as high computational costs. This paper proposes the Multi-input and Multi-scale Efficient Transformer (MMET), a novel framework designed to address the above challenges. MMET decouples mesh and query points as two sequences and feeds them into the encoder and decoder, respectively, and uses a Gated Condition Embedding (GCE) layer to embed input variables or functions with varying dimensions, enabling effective solutions for multi-scale and multi-input problems. Additionally, a Hilbert curve-based reserialization and patch embedding mechanism decrease the input length. This significantly reduces the computational cost when dealing with large-scale geometric models. These innovations enable efficient representations and support multi-scale resolution queries for large-scale and multi-input PDE problems. Experimental evaluations on diverse benchmarks spanning different physical fields demonstrate that MMET outperforms SOTA methods in both accuracy and computational efficiency. This work highlights the potential of MMET as a robust and scalable solution for real-time PDE solving in engineering and physics-based applications, paving the way for future explorations into pre-trained large-scale models in specific domains. This work is open-sourced at https://github.com/YichenLuo-0/MMET.
Authors: Shuyin Xia, Yifan Wang, Lifeng Shen, Guoyin Wang
Abstract: Most existing multi-kernel clustering algorithms, such as multi-kernel K-means, often struggle with computational efficiency and robustness when faced with complex data distributions. These challenges stem from their dependence on point-to-point relationships for optimization, which can lead to difficulty in accurately capturing data sets' inherent structure and diversity. Additionally, the intricate interplay between multiple kernels in such algorithms can further exacerbate these issues, effectively impacting their ability to cluster data points in high-dimensional spaces. In this paper, we leverage granular-ball computing to improve the multi-kernel clustering framework. The core of granular-ball computing is to adaptively fit data distribution by balls from coarse to acceptable levels. Each ball can enclose data points based on a density consistency measurement. Such ball-based data description thus improves the computational efficiency and the robustness to unknown noises. Specifically, based on granular-ball representations, we introduce the granular-ball kernel (GBK) and its corresponding granular-ball multi-kernel K-means framework (GB-MKKM) for efficient clustering. Using granular-ball relationships in multiple kernel spaces, the proposed GB-MKKM framework shows its superiority in efficiency and clustering performance in the empirical evaluation of various clustering tasks.
Authors: Shili Wu, Yizhao Jin, Puhua Niu, Aniruddha Datta, Sean B. Andersson
Abstract: Behavior Cloning (BC) is an effective imitation learning technique and has even been adopted in some safety-critical domains such as autonomous vehicles. BC trains a policy to mimic the behavior of an expert by using a dataset composed of only state-action pairs demonstrated by the expert, without any additional interaction with the environment. However, During deployment, the policy observations may contain measurement errors or adversarial disturbances. Since the observations may deviate from the true states, they can mislead the agent into making sub-optimal actions. In this work, we use a global Lipschitz regularization approach to enhance the robustness of the learned policy network. We then show that the resulting global Lipschitz property provides a robustness certificate to the policy with respect to different bounded norm perturbations. Then, we propose a way to construct a Lipschitz neural network that ensures the policy robustness. We empirically validate our theory across various environments in Gymnasium. Keywords: Robust Reinforcement Learning; Behavior Cloning; Lipschitz Neural Network
Authors: Youkang Wang, Jian Wang, Rubing Chen, Xiao-Yong Wei
Abstract: Inference-time scaling has emerged as a powerful technique for enhancing the reasoning performance of Large Language Models (LLMs). However, existing approaches often rely on heuristic strategies for parallel sampling, lacking a principled foundation. To address this gap, we propose a probabilistic framework that formalizes the optimality of inference-time scaling under the assumption that parallel samples are independently and identically distributed (i.i.d.), and where the Best-of-N selection strategy follows a probability distribution that can be estimated. Within this framework, we derive a theoretical lower bound on the required number of samples to achieve a target performance level, providing the first principled guidance for compute-efficient scaling. Leveraging this insight, we develop OptScale, a practical algorithm that dynamically determines the optimal number of sampled responses. OptScale employs a language model-based predictor to estimate probabilistic prior parameters, enabling the decision of the minimal number of samples needed that satisfy predefined performance thresholds and confidence levels. Extensive experiments on mathematical reasoning benchmarks (including MATH-500, GSM8K, AIME, and AMC) demonstrate that OptScale significantly reduces sampling overhead while remaining better or on par with state-of-the-art reasoning performance. Our work offers both a theoretical foundation and a practical solution for principled inference-time scaling, addressing a critical gap in the efficient deployment of LLMs for complex reasoning. The source code is publicly available at https://github.com/Albertwyk/OptScale.
Authors: Xingke Yang, Liang Li, Zhiyi Wan, Sicong Li, Xiaoqi Qi, Jiang Liu, Tomoaki Ohtsuki, Xin Fu, Miao Pan
Abstract: There is a huge gap between numerous intriguing applications fostered by on-device large language model (LLM) fine-tuning (FT) from fresh mobile data and the limited resources of a mobile device. While existing server-assisted methods (e.g., split learning or side-tuning) may enable LLM FT on the local mobile device, they suffer from heavy communication burdens of activation transmissions, and may disclose data and labels to the server. To address those issues, we develop PAE MobiLLM, a a privacy-aware and efficient LLM FT method which can be deployed on the mobile device via server-assisted additive side-tuning. To further accelerate FT convergence and improve computing efficiency, PAE MobiLLM integrates activation caching on the server side, which allows the server to reuse historical activations and saves the mobile device from repeatedly computing forward passes for the recurring data samples. Besides, to reduce communication cost, PAE MobiLLM develops an activation shortcut that transmits only the token involved in the loss calculation instead of full activation matrices to guide the side network tuning. Last but not least, PAE MobiLLM introduces the additive adapter side-network design which makes the server train the adapter modules based on device-defined prediction differences rather than raw ground-truth labels. In this way, the server can only assist device-defined side-network computing, and learn nothing about data and labels. Extensive experimental results demonstrate PAE MobiLLM's superiority.
Authors: Huibo Xu, Runlong Yu, Likang Wu, Xianquan Wang, Qi Liu
Abstract: Existing generative models for time series forecasting often transform simple priors (typically Gaussian) into complex data distributions. However, their sampling initialization, independent of historical data, hinders the capture of temporal dependencies, limiting predictive accuracy. They also treat residuals merely as optimization targets, ignoring that residuals often exhibit meaningful patterns like systematic biases or nontrivial distributional structures. To address these, we propose Conditional Guided Flow Matching (CGFM), a novel model-agnostic framework that extends flow matching by integrating outputs from an auxiliary predictive model. This enables learning from the probabilistic structure of prediction residuals, leveraging the auxiliary model's prediction distribution as a source to reduce learning difficulty and refine forecasts. CGFM incorporates historical data as both conditions and guidance, uses two-sided conditional paths (with source and target conditioned on the same history), and employs affine paths to expand the path space, avoiding path crossing without complex mechanisms, preserving temporal consistency, and strengthening distribution alignment. Experiments across datasets and baselines show CGFM consistently outperforms state-of-the-art models, advancing forecasting.
Authors: Shim Soon Yong
Abstract: We introduce a novel classification framework, ZClassifier, that replaces conventional deterministic logits with diagonal Gaussian-distributed logits. Our method simultaneously addresses temperature scaling and manifold approximation by minimizing the KL divergence between the predicted Gaussian distributions and a unit isotropic Gaussian. This unifies uncertainty calibration and latent control in a principled probabilistic manner, enabling a natural interpretation of class confidence and geometric consistency. Experiments on CIFAR-10 and CIFAR-100 demonstrate that ZClassifier improves over softmax classifiers in robustness, calibration, and latent separation, with consistent benefits across small-scale and large-scale classification settings.
Authors: Andris Ambainis, Joao F. Doriguello, Debbie Lim
Abstract: We propose novel classical and quantum online algorithms for learning finite-horizon and infinite-horizon average-reward Markov Decision Processes (MDPs). Our algorithms are based on a hybrid exploration-generative reinforcement learning (RL) model wherein the agent can, from time to time, freely interact with the environment in a generative sampling fashion, i.e., by having access to a "simulator". By employing known classical and new quantum algorithms for approximating optimal policies under a generative model within our learning algorithms, we show that it is possible to avoid several paradigms from RL like "optimism in the face of uncertainty" and "posterior sampling" and instead compute and use optimal policies directly, which yields better regret bounds compared to previous works. For finite-horizon MDPs, our quantum algorithms obtain regret bounds which only depend logarithmically on the number of time steps $T$, thus breaking the $O(\sqrt{T})$ classical barrier. This matches the time dependence of the prior quantum works of Ganguly et al. (arXiv'23) and Zhong et al. (ICML'24), but with improved dependence on other parameters like state space size $S$ and action space size $A$. For infinite-horizon MDPs, our classical and quantum bounds still maintain the $O(\sqrt{T})$ dependence but with better $S$ and $A$ factors. Nonetheless, we propose a novel measure of regret for infinite-horizon MDPs with respect to which our quantum algorithms have $\operatorname{poly}\log{T}$ regret, exponentially better compared to classical algorithms. Finally, we generalise all of our results to compact state spaces.
Authors: Kexin Zhao, Bo Wang, Cuiying Zhao, Tongyao Wan
Abstract: Optimizing credit limits, interest rates, and loan terms is crucial for managing borrower risk and lifetime value (LTV) in personal loan platform. However, counterfactual estimation of these continuous, multi-dimensional treatments faces significant challenges: randomized trials are often prohibited by risk controls and long repayment cycles, forcing reliance on biased observational data. Existing causal methods primarily handle binary/discrete treatments and struggle with continuous, multi-dimensional settings. Furthermore, financial domain knowledge mandates provably monotonic treatment-outcome relationships (e.g., risk increases with credit limit).To address these gaps, we propose Multi-Treatment-DML, a novel framework leveraging Double Machine Learning (DML) to: (i) debias observational data for causal effect estimation; (ii) handle arbitrary-dimensional continuous treatments; and (iii) enforce monotonic constraints between treatments and outcomes, guaranteeing adherence to domain requirements.Extensive experiments on public benchmarks and real-world industrial datasets demonstrate the effectiveness of our approach. Furthermore, online A/B testing conducted on a realworld personal loan platform, confirms the practical superiority of Multi-Treatment-DML in real-world loan operations.
Authors: Biyi Fang, Jean Utke, Truong Vo, Diego Klabjan
Abstract: Transformer architectures dominate modern NLP but often demand heavy computational resources and intricate hyperparameter tuning. To mitigate these challenges, we propose a novel framework, BoostTransformer, that augments transformers with boosting principles through subgrid token selection and importance-weighted sampling. Our method incorporates a least square boosting objective directly into the transformer pipeline, enabling more efficient training and improved performance. Across multiple fine-grained text classification benchmarks, BoostTransformer demonstrates both faster convergence and higher accuracy, surpassing standard transformers while minimizing architectural search overhead.
Authors: Mengting Pan, Fan Li, Xiaoyang Wang, Wenjie Zhang, Xuemin Lin
Abstract: Contrastive learning (CL) has become a dominant paradigm for self-supervised hypergraph learning, enabling effective training without costly labels. However, node entities in real-world hypergraphs are often associated with rich textual information, which is overlooked in prior works. Directly applying existing CL-based methods to such text-attributed hypergraphs (TAHGs) leads to three key limitations: (1) The common use of graph-agnostic text encoders overlooks the correlations between textual content and hypergraph topology, resulting in suboptimal representations. (2) Their reliance on random data augmentations introduces noise and weakens the contrastive objective. (3) The primary focus on node- and hyperedge-level contrastive signals limits the ability to capture long-range dependencies, which is essential for expressive representation learning. Although HyperBERT pioneers CL on TAHGs, its co-training paradigm suffers from poor scalability. To fill the research gap, we introduce HiTeC, a two-stage hierarchical contrastive learning framework with semantic-aware augmentation for scalable and effective self-supervised learning on TAHGs. In the first stage, we pre-train the text encoder with a structure-aware contrastive objective to overcome the graph-agnostic nature of conventional methods. In the second stage, we introduce two semantic-aware augmentation strategies, including prompt-enhanced text augmentation and semantic-aware hyperedge drop, to facilitate informative view generation. Furthermore, we propose a multi-scale contrastive loss that extends existing objectives with an $s$-walk-based subgraph-level contrast to better capture long-range dependencies. By decoupling text encoder pretraining from hypergraph contrastive learning, this two-stage design enhances scalability without compromising representation quality. Extensive experiments confirm the effectiveness of HiTeC.
Authors: Zahra Ebrahimi, Raheleh Salehi, Nassir Navab, Carsten Marr, Ario Sadafi
Abstract: The dynamic environment of laboratories and clinics, with streams of data arriving on a daily basis, requires regular updates of trained machine learning models for consistent performance. Continual learning is supposed to help train models without catastrophic forgetting. However, state-of-the-art methods are ineffective for multiple instance learning (MIL), which is often used in single-cell-based hematologic disease diagnosis (e.g., leukemia detection). Here, we propose the first continual learning method tailored specifically to MIL. Our method is rehearsal-based over a selection of single instances from various bags. We use a combination of the instance attention score and distance from the bag mean and class mean vectors to carefully select which samples and instances to store in exemplary sets from previous tasks, preserving the diversity of the data. Using the real-world input of one month of data from a leukemia laboratory, we study the effectiveness of our approach in a class incremental scenario, comparing it to well-known continual learning methods. We show that our method considerably outperforms state-of-the-art methods, providing the first continual learning approach for MIL. This enables the adaptation of models to shifting data distributions over time, such as those caused by changes in disease occurrence or underlying genetic alterations.
Authors: Jie Xiao, Changyuan Fan, Qingnan Ren, Alfred Long, Yuchen Zhang, Rymon Yu, Eric Yang, Lynn Ai, Shaoduo Gan
Abstract: Modern RL-based post-training for large language models (LLMs) co-locate trajectory sampling and policy optimisation on the same GPU cluster, forcing the system to switch between inference and training workloads. This serial context switching violates the single-program-multiple-data (SPMD) assumption underlying today's distributed training systems. We present Echo, the RL system that cleanly decouples these two phases across heterogeneous "inference" and "training" swarms while preserving statistical efficiency. Echo introduces two lightweight synchronization protocols: a sequential pull mode that refreshes policy weights according to API call for minimal bias, and an asynchronous push-pull mode that streams version-tagged rollouts through a replay buffer to maximise hardware utilisation. Training four representative RL workloads with Qwen3-4B, Qwen2.5-7B, Qwen3-30B-A3B-Thinking-2507 and Qwen3-32B on a geographically distributed cluster, Echo matches a fully co-located Verl baseline in convergence speed and final reward while off-loading trajectory generation to commodity edge hardware. These promising results demonstrate that large-scale RL for LLMs could achieve datacentre-grade performance using decentralised, heterogeneous resources.
Authors: Jeroen F. Uleman, Loes Crielaard, Leonie K. Elsenburg, Guido A. Veldhuis, Karien Stronks, Naja Hulvej Rod, Rick Quax, V\'itor V. Vasconcelos
Abstract: Causal loop diagrams (CLDs) are widely used in health and environmental research to represent hypothesized causal structures underlying complex problems. However, as qualitative and static representations, CLDs are limited in their ability to support dynamic analysis and inform intervention strategies. Additionally, quantitative CLD analysis methods like network centrality analysis often lead to false inference. We propose Diagrams-to-Dynamics (D2D), a method for converting CLDs into exploratory system dynamics models (SDMs) in the absence of empirical data. With minimal user input - following a protocol to label variables as stocks, flows or auxiliaries, and constants - D2D leverages the structural information already encoded in CLDs, namely, link existence and polarity, to simulate hypothetical interventions and explore potential leverage points under uncertainty. Results suggest that D2D helps distinguish between high- and low-ranked leverage points. We compare D2D to a data-driven SDM constructed from the same CLD and variable labels. D2D showed greater consistency with the data-driven model than network centrality analysis, while providing uncertainty estimates and guidance for future data collection. The method is implemented in an open-source Python package and a web-based application to support further testing and lower the barrier to dynamic modeling for researchers working with CLDs. We expect additional validation will further establish the approach's utility across a broad range of cases and domains.
Authors: Anurag Tripathi, Vaibhav Patle, Abhinav Jain, Ayush Pundir, Sairam Menon, Ajeet Kumar Singh, Dorien Herremans
Abstract: Text-to-SQL bridges the gap between natural language and structured database language, thus allowing non-technical users to easily query databases. Traditional approaches model text-to-SQL as a direct translation task, where a given Natural Language Query (NLQ) is mapped to an SQL command. Recent advances in large language models (LLMs) have significantly improved translation accuracy, however, these methods all require that the target database is pre-specified. This becomes problematic in scenarios with multiple extensive databases, where identifying the correct database becomes a crucial yet overlooked step. In this paper, we propose a three-stage end-to-end text-to-SQL framework to identify the user's intended database before generating SQL queries. Our approach leverages LLMs and prompt engineering to extract implicit information from natural language queries (NLQs) in the form of a ruleset. We then train a large db\_id prediction model, which includes a RoBERTa-based finetuned encoder, to predict the correct Database identifier (db\_id) based on both the NLQ and the LLM-generated rules. Finally, we refine the generated SQL by using critic agents to correct errors. Experimental results demonstrate that our framework outperforms the current state-of-the-art models in both database intent prediction and SQL generation accuracy.
Authors: Wooyong Jung, Sola Kim, Dongwook Kim, Maryam Tabar, Dongwon Lee
Abstract: Homelessness in the United States has surged to levels unseen since the Great Depression. However, existing methods for monitoring it, such as point-in-time (PIT) counts, have limitations in terms of frequency, consistency, and spatial detail. This study proposes a new approach using publicly available, crowdsourced data, specifically 311 Service Calls and street-level imagery, to track and forecast homeless tent trends in San Francisco. Our predictive model captures fine-grained daily and neighborhood-level variations, uncovering patterns that traditional counts often overlook, such as rapid fluctuations during the COVID-19 pandemic and spatial shifts in tent locations over time. By providing more timely, localized, and cost-effective information, this approach serves as a valuable tool for guiding policy responses and evaluating interventions aimed at reducing unsheltered homelessness.
Authors: Qingguo Wang
Abstract: Measles, a highly contagious disease declared eliminated in the United States in 2000 after decades of successful vaccination campaigns, resurged in 2025, with 1,356 confirmed cases reported as of August 5, 2025. Given its rapid spread among susceptible individuals, fast and reliable diagnostic systems are critical for early prevention and containment. In this work, we applied transfer learning to fine-tune a pretrained Data-efficient Image Transformer (DeiT) model for distinguishing measles rashes from other skin conditions. Trained on a diverse, curated skin rash image dataset, the DeiT model achieved a median classification accuracy of 96.38%, precision of 96.24%, recall of 96.38%, and an F1-score of 96.23%, demonstrating high effectiveness in accurate detection to aid outbreak control. We also compared the DeiT model with a convolutional neural network, ResNet-50, and discussed the directions for future research.
Authors: Elie Azeraf, Emmanuel Monfrini, Emmanuel Vignon, Wojciech Pieczynski
Abstract: Natural Language Processing (NLP) models' current trend consists of using increasingly more extra-data to build the best models as possible. It implies more expensive computational costs and training time, difficulties for deployment, and worries about these models' carbon footprint reveal a critical problem in the future. Against this trend, our goal is to develop NLP models requiring no extra-data and minimizing training time. To do so, in this paper, we explore Markov chain models, Hidden Markov Chain (HMC) and Pairwise Markov Chain (PMC), for NLP segmentation tasks. We apply these models for three classic applications: POS Tagging, Named-Entity-Recognition, and Chunking. We develop an original method to adapt these models for text segmentation's specific challenges to obtain relevant performances with very short training and execution times. PMC achieves equivalent results to those obtained by Conditional Random Fields (CRF), one of the most applied models for these tasks when no extra-data are used. Moreover, PMC has training times 30 times shorter than the CRF ones, which validates this model given our objectives.
Authors: Xinyun Chen, Guiyu Hong, Yunan Liu
Abstract: We investigate an optimization problem in a queueing system where the service provider selects the optimal service fee p and service capacity \mu to maximize the cumulative expected profit (the service revenue minus the capacity cost and delay penalty). The conventional predict-then-optimize (PTO) approach takes two steps: first, it estimates the model parameters (e.g., arrival rate and service-time distribution) from data; second, it optimizes a model based on the estimated parameters. A major drawback of PTO is that its solution accuracy can often be highly sensitive to the parameter estimation errors because PTO is unable to properly link these errors (step 1) to the quality of the optimized solutions (step 2). To remedy this issue, we develop an online learning framework that automatically incorporates the aforementioned parameter estimation errors in the solution prescription process; it is an integrated method that can "learn" the optimal solution without needing to set up the parameter estimation as a separate step as in PTO. Effectiveness of our online learning approach is substantiated by (i) theoretical results including the algorithm convergence and analysis of the regret ("cost" to pay over time for the algorithm to learn the optimal policy), and (ii) engineering confirmation via simulation experiments of a variety of representative examples. We also provide careful comparisons for PTO and the online learning method.
Authors: Jongyeong Lee, Junya Honda, Masashi Sugiyama
Abstract: This paper studies the fixed-confidence best arm identification (BAI) problem in the bandit framework in the canonical single-parameter exponential models. For this problem, many policies have been proposed, but most of them require solving an optimization problem at every round and/or are forced to explore an arm at least a certain number of times except those restricted to the Gaussian model. To address these limitations, we propose a novel policy that combines Thompson sampling with a computationally efficient approach known as the best challenger rule. While Thompson sampling was originally considered for maximizing the cumulative reward, we demonstrate that it can be used to naturally explore arms in BAI without forcing it. We show that our policy is asymptotically optimal for any two-armed bandit problems and achieves near optimality for general $K$-armed bandit problems for $K\geq 3$. Nevertheless, in numerical experiments, our policy shows competitive performance compared to asymptotically optimal policies in terms of sample complexity while requiring less computation cost. In addition, we highlight the advantages of our policy by comparing it to the concept of $\beta$-optimality, a relaxed notion of asymptotic optimality commonly considered in the analysis of a class of policies including the proposed one.
Authors: Md Rakibul Hasan, Md Zakir Hossain, Shreya Ghosh, Aneesh Krishna, Tom Gedeon
Abstract: Empathy indicates an individual's ability to understand others. Over the past few years, empathy has drawn attention from various disciplines, including but not limited to Affective Computing, Cognitive Science, and Psychology. Detecting empathy has potential applications in society, healthcare and education. Despite being a broad and overlapping topic, the avenue of empathy detection leveraging Machine Learning remains underexplored from a systematic literature review perspective. We collected 849 papers from 10 well-known academic databases, systematically screened them and analysed the final 82 papers. Our analyses reveal several prominent task formulations - including empathy on localised utterances or overall expressions, unidirectional or parallel empathy, and emotional contagion - in monadic, dyadic and group interactions. Empathy detection methods are summarised based on four input modalities - text, audiovisual, audio and physiological signals - thereby presenting modality-specific network architecture design protocols. We discuss challenges, research gaps and potential applications in the Affective Computing-based empathy domain, which can facilitate new avenues of exploration. We further enlist the public availability of datasets and codes. This paper, therefore, provides a structured overview of recent advancements and remaining challenges towards developing a robust empathy detection system that could meaningfully contribute to enhancing human well-being.
Authors: Pourya Behmandpoor, Mark Eisen, Panagiotis Patrinos, Marc Moonen
Abstract: Deep learning (DL) based resource allocation (RA) has recently gained significant attention due to its performance efficiency. However, most related studies assume an ideal case where the number of users and their utility demands, e.g., data rate constraints, are fixed, and the designed DL-based RA scheme exploits a policy trained only for these fixed parameters. Consequently, computationally complex policy retraining is required whenever these parameters change. In this paper, we introduce a DL-based resource allocator (ALCOR) that allows users to adjust their utility demands freely, such as based on their application layer requirements. ALCOR employs deep neural networks (DNNs) as the policy in a time-sharing problem. The underlying optimization algorithm iteratively optimizes the on-off status of users to satisfy their utility demands in expectation. The policy performs unconstrained RA (URA) -- RA without considering user utility demands -- among active users to maximize the sum utility (SU) at each time instant. Depending on the chosen URA scheme, ALCOR can perform RA in either a centralized or distributed scenario. The derived convergence analyses provide theoretical guarantees for ALCOR's convergence, and numerical experiments corroborate its effectiveness compared to meta-learning and reinforcement learning approaches.
Authors: S. Dubey, T. E. Browder, S. Kohani, R. Mandal, A. Sibidanov, R. Sinha
Abstract: We report on a novel application of computer vision techniques to extract beyond the Standard Model parameters directly from high energy physics flavor data. We propose a novel data representation that transforms the angular and kinematic distributions into ``quasi-images", which are used to train a convolutional neural network to perform regression tasks, similar to fitting. As a proof-of-concept, we train a 34-layer Residual Neural Network to regress on these images and determine information about the Wilson Coefficient $C_{9}$ in Monte Carlo simulations of $B^0 \rightarrow K^{*0}\mu^{+}\mu^{-}$ decays. The method described here can be generalized and may find applicability across a variety of experiments.
Authors: Feiyang Wu, Xavier Nal, Jaehwi Jang, Wei Zhu, Zhaoyuan Gu, Anqi Wu, Ye Zhao
Abstract: Humanoid robots promise transformative capabilities for industrial and service applications. While recent advances in Reinforcement Learning (RL) yield impressive results in locomotion, manipulation, and navigation, the proposed methods typically require enormous simulation samples to account for real-world variability. This work proposes a novel one-stage training framework-Learn to Teach (L2T)-which unifies teacher and student policy learning. Our approach recycles simulator samples and synchronizes the learning trajectories through shared dynamics, significantly reducing sample complexities and training time while achieving state-of-the-art performance. Furthermore, we validate the RL variant (L2T-RL) through extensive simulations and hardware tests on the Digit robot, demonstrating zero-shot sim-to-real transfer and robust performance over 12+ challenging terrains without depth estimation modules.
Authors: Yuxuan Li, Xiang Li, Weijie Li, Qibin Hou, Li Liu, Ming-Ming Cheng, Jian Yang
Abstract: Synthetic Aperture Radar (SAR) object detection has gained significant attention recently due to its irreplaceable all-weather imaging capabilities. However, this research field suffers from both limited public datasets (mostly comprising <2K images with only mono-category objects) and inaccessible source code. To tackle these challenges, we establish a new benchmark dataset and an open-source method for large-scale SAR object detection. Our dataset, SARDet-100K, is a result of intense surveying, collecting, and standardizing 10 existing SAR detection datasets, providing a large-scale and diverse dataset for research purposes. To the best of our knowledge, SARDet-100K is the first COCO-level large-scale multi-class SAR object detection dataset ever created. With this high-quality dataset, we conducted comprehensive experiments and uncovered a crucial challenge in SAR object detection: the substantial disparities between the pretraining on RGB datasets and finetuning on SAR datasets in terms of both data domain and model structure. To bridge these gaps, we propose a novel Multi-Stage with Filter Augmentation (MSFA) pretraining framework that tackles the problems from the perspective of data input, domain transition, and model migration. The proposed MSFA method significantly enhances the performance of SAR object detection models while demonstrating exceptional generalizability and flexibility across diverse models. This work aims to pave the way for further advancements in SAR object detection. The dataset and code is available at https://github.com/zcablii/SARDet_100K.
Authors: Isma\"el Castillo, Alice L'Huillier, Kolyan Ray, Luke Travis
Abstract: We propose a scalable variational Bayes method for statistical inference for a single or low-dimensional subset of the coordinates of a high-dimensional parameter in sparse linear regression. Our approach relies on assigning a mean-field approximation to the nuisance coordinates and carefully modelling the conditional distribution of the target given the nuisance. This requires only a preprocessing step and preserves the computational advantages of mean-field variational Bayes, while ensuring accurate and reliable inference for the target parameter, including for uncertainty quantification. We investigate the numerical performance of our algorithm, showing that it performs competitively with existing methods. We further establish accompanying theoretical guarantees for estimation and uncertainty quantification in the form of a Bernstein--von Mises theorem.
Authors: Rishabh Anand, Chaitanya K. Joshi, Alex Morehead, Arian R. Jamasb, Charles Harris, Simon V. Mathis, Kieran Didi, Rex Ying, Bryan Hooi, Pietro Li\`o
Abstract: We introduce RNA-FrameFlow, the first generative model for 3D RNA backbone design. We build upon SE(3) flow matching for protein backbone generation and establish protocols for data preparation and evaluation to address unique challenges posed by RNA modeling. We formulate RNA structures as a set of rigid-body frames and associated loss functions which account for larger, more conformationally flexible RNA backbones (13 atoms per nucleotide) vs. proteins (4 atoms per residue). Toward tackling the lack of diversity in 3D RNA datasets, we explore training with structural clustering and cropping augmentations. Additionally, we define a suite of evaluation metrics to measure whether the generated RNA structures are globally self-consistent (via inverse folding followed by forward folding) and locally recover RNA-specific structural descriptors. The most performant version of RNA-FrameFlow generates locally realistic RNA backbones of 40-150 nucleotides, over 40% of which pass our validity criteria as measured by a self-consistency TM-score >= 0.45, at which two RNAs have the same global fold. Open-source code: https://github.com/rish-16/rna-backbone-design
Authors: Walter Laurito, Benjamin Davis, Peli Grietzer, Tom\'a\v{s} Gaven\v{c}iak, Ada B\"ohm, Jan Kulveit
Abstract: Are large language models (LLMs) biased in favor of communications produced by LLMs, leading to possible antihuman discrimination? Using a classical experimental design inspired by employment discrimination studies, we tested widely used LLMs, including GPT-3.5, GPT-4 and a selection of recent open-weight models in binary choice scenarios. These involved LLM-based assistants selecting between goods (the goods we study include consumer products, academic papers, and film-viewings) described either by humans or LLMs. Our results show a consistent tendency for LLM-based AIs to prefer LLM-presented options. This suggests the possibility of future AI systems implicitly discriminating against humans as a class, giving AI agents and AI-assisted humans an unfair advantage.
Authors: Jakub Adamczyk, Piotr Ludynia
Abstract: In this work, we present scikit-fingerprints, a Python package for computation of molecular fingerprints for applications in chemoinformatics. Our library offers an industry-standard scikit-learn interface, allowing intuitive usage and easy integration with machine learning pipelines. It is also highly optimized, featuring parallel computation that enables efficient processing of large molecular datasets. Currently, scikit-fingerprints stands as the most feature-rich library in the open source Python ecosystem, offering over 30 molecular fingerprints. Our library simplifies chemoinformatics tasks based on molecular fingerprints, including molecular property prediction and virtual screening. It is also flexible, highly efficient, and fully open source.
Authors: Cho Yang
Abstract: This paper proposes a novel model inspired by CycleGAN: FQGA-single to produce high quality medical synthetic CT (sCT) generated images more efficiently. Evaluations were done on the SynthRAD Grand Challenge dataset with the CycleGAN model used for benchmarking and for comparing the quality of CBCT-to-sCT generated images from both a quantitative and qualitative perspective. Finally, this paper also explores ideas from the paper "One Epoch Is All You Need" to compare models trained on a single epoch versus multiple epochs. Astonishing results from FQGA-single were obtained during this exploratory experiment, which show that the performance of FQGA-single when trained on a single epoch surpasses itself when trained on multiple epochs. More surprising is that its performance also surpasses CycleGAN's multiple-epoch and single-epoch models, and even a modified version of CycleGAN.
Authors: Ningyuan Xi, Yetao Wu, Kun Fan, Teng Chen, Qingqing Gu, Luo Ji
Abstract: Large Language Models (LLM) often need to be Continual Pre-Trained (CPT) to obtain unfamiliar language skills or adapt to new domains. The huge training cost of CPT often asks for cautious choice of key hyper-parameters such as the mixture ratio of extra language or domain corpus. However, there is no systematic study that bridges the gap between the optimal mixture ratio and the actual model performance, and the gap between experimental scaling law and the actual deployment in the full model size. In this paper, we perform CPT on Llama-3 8B and 70B to enhance its Chinese ability. We study the optimal correlation between the Additional Language Mixture Ratio (ALMR) and the Learning Rate (LR) on the 8B size which directly indicates the optimal experimental setup. By thorough choice of hyper-parameter, and subsequent fine-tuning, the model capability is improved not only on the Chinese-related benchmark but also in some specific domains including math, coding, and emotional intelligence. We deploy the final 70B version of LLM on a real-life chat system which obtains satisfying performance.
Authors: Sam Gijsen, Kerstin Ritter
Abstract: Multimodal language modeling has enabled breakthroughs for representation learning, yet remains unexplored in the realm of functional brain data for clinical phenotyping. This paper pioneers EEG-language models (ELMs) trained on clinical reports and 15000 EEGs. We propose to combine multimodal alignment in this novel domain with timeseries cropping and text segmentation, enabling an extension based on multiple instance learning to alleviate misalignment between irrelevant EEG or text segments. Our multimodal models significantly improve over EEG-only models across four clinical evaluations and for the first time enable zero-shot classification as well as retrieval of both neural signals and reports. In sum, these results highlight the potential of ELMs, representing significant progress for clinical applications.
Authors: Mohammad Mehdi Rastikerdar, Jin Huang, Hui Guan, Deepak Ganesan
Abstract: Resource-constrained IoT devices increasingly rely on deep learning models, however, these models experience significant accuracy drops due to domain shifts when encountering variations in lighting, weather, and seasonal conditions. While cloud-based retraining can address this issue, many IoT deployments operate with limited connectivity and energy constraints, making traditional fine-tuning approaches impractical. We explore this challenge through the lens of wildlife ecology, where camera traps must maintain accurate species classification across changing seasons, weather, and habitats without reliable connectivity. We introduce WildFit, an autonomous in-situ adaptation framework that leverages the key insight that background scenes change more frequently than the visual characteristics of monitored species. WildFit combines background-aware synthesis to generate training samples on-device with drift-aware fine-tuning that triggers model updates only when necessary to conserve resources. Our background-aware synthesis surpasses efficient baselines by 7.3\% and diffusion models by 3.0\% while being orders of magnitude faster, our drift-aware fine-tuning achieves Pareto optimality with 50\% fewer updates and 1.5\% higher accuracy, and the end-to-end system outperforms domain adaptation approaches by 20--35%\% while consuming only 11.2 Wh over 37 days -- enabling battery-powered deployment.
Authors: Bingzhi Zhang, Junyu Liu, Liang Jiang, Quntao Zhuang
Abstract: Quantum neural networks, parameterized quantum circuits optimized under a specific cost function, provide a paradigm for achieving near-term quantum advantage in quantum information processing. Understanding QNN training dynamics is crucial for optimizing their performance, however, the role of quantum data in training for supervised learning such as classification and regression remains unclear. We reveal a quantum-data-driven dynamical transition where the target values and data determine the convergence of the training. Through analytical classification over the fixed points of the dynamical equation, we reveal a comprehensive `phase diagram' featuring seven distinct dynamics originating from a bifurcation with multiple codimension. Perturbative analyses identify both exponential and polynomial convergence class. We provide a non-perturbative theory to explain the transition via generalized restricted Haar ensemble. The analytical results are confirmed with numerical simulations and experimentation on IBM quantum devices. Our findings provide guidance on constructing the cost function to accelerate convergence in QNN training.
Authors: Xiaojian Yuan, Tianyu Pang, Chao Du, Kejiang Chen, Weiming Zhang, Min Lin
Abstract: Large language models (LLMs) may memorize sensitive or copyrighted content, raising privacy and legal concerns. Due to the high cost of retraining from scratch, researchers attempt to employ machine unlearning to remove specific content from LLMs while preserving the overall performance. In this paper, we discuss several issues in machine unlearning for LLMs and provide our insights on possible approaches. To address the issue of inadequate evaluation of model outputs after unlearning, we introduce three additional metrics to evaluate token diversity, sentence semantics, and factual correctness. We then categorize unlearning methods into untargeted and targeted, and discuss their issues respectively. Specifically, the behavior that untargeted unlearning attempts to approximate is unpredictable and may involve hallucinations, and existing regularization is insufficient for targeted unlearning. To alleviate these issues, we propose using the objective of maximizing entropy (ME) for untargeted unlearning and incorporate answer preservation (AP) loss as regularization for targeted unlearning. Experimental results across three scenarios, i.e., fictitious unlearning, continual unlearning, and real-world unlearning, demonstrate the effectiveness of our approaches. The code is available at https://github.com/sail-sg/closer-look-LLM-unlearning.
URLs: https://github.com/sail-sg/closer-look-LLM-unlearning.
Authors: Yuxuan Sun, Ruikang Liu, Haoli Bai, Han Bao, Kang Zhao, Yuening Li, Jiaxin Hu, Xianzhi Yu, Lu Hou, Chun Yuan, Xin Jiang, Wulong Liu, Jun Yao
Abstract: Recently, quantization has been widely used for the compression and acceleration of large language models (LLMs). Due to the outliers in LLMs, it is crucial to flatten weights and activations to minimize quantization error with equally spaced quantization points. Prior research explores various pre-quantization transformations to suppress outliers, such as per-channel scaling and Hadamard transformation. However, we observe that these transformed weights and activations can still exhibit steep and dispersed distributions. In this paper, we propose FlatQuant (Fast and Learnable Affine Transformation), a new post-training quantization approach that enhances the flatness of weights and activations. Our approach identifies optimal affine transformations for each linear layer, calibrated in hours via a lightweight objective. To reduce runtime overhead of affine transformation, we apply Kronecker product with two lightweight matrices, and fuse all operations in FlatQuant into a single kernel. Extensive experiments demonstrate that FlatQuant establishes a new state-of-the-art benchmark for quantization. For example, it achieves less than 1\% accuracy drop for W4A4 quantization on the LLaMA-3-70B model, surpassing SpinQuant by 7.5\%. Additionally, it provides up to 2.3x prefill speedup and 1.7x decoding speedup compared to the FP16 model. Code is available at: https://github.com/ruikangliu/FlatQuant.
Authors: Hongcheng Gao, Tianyu Pang, Chao Du, Taihang Hu, Zhijie Deng, Min Lin
Abstract: With the rapid progress of diffusion-based content generation, significant efforts are being made to unlearn harmful or copyrighted concepts from pretrained diffusion models (DMs) to prevent potential model misuse. However, it is observed that even when DMs are properly unlearned before release, malicious finetuning can compromise this process, causing DMs to relearn the unlearned concepts. This occurs partly because certain benign concepts (e.g., "skin") retained in DMs are related to the unlearned ones (e.g., "nudity"), facilitating their relearning via finetuning. To address this, we propose meta-unlearning on DMs. Intuitively, a meta-unlearned DM should behave like an unlearned DM when used as is; moreover, if the meta-unlearned DM undergoes malicious finetuning on unlearned concepts, the related benign concepts retained within it will be triggered to self-destruct, hindering the relearning of unlearned concepts. Our meta-unlearning framework is compatible with most existing unlearning methods, requiring only the addition of an easy-to-implement meta objective. We validate our approach through empirical experiments on meta-unlearning concepts from Stable Diffusion models (SD-v1-4 and SDXL), supported by extensive ablation studies. Our code is available at https://github.com/sail-sg/Meta-Unlearning.
Authors: Runshi Tang, Tamara Kolda, Anru R. Zhang
Abstract: This paper presents a canonical polyadic (CP) tensor decomposition that addresses unaligned observations. The mode with unaligned observations is represented using functions in a reproducing kernel Hilbert space (RKHS). We introduce a versatile loss function that effectively accounts for various types of data, including binary, integer-valued, and positive-valued types. Additionally, we propose an optimization algorithm for computing tensor decompositions with unaligned observations, along with a stochastic gradient method to enhance computational efficiency. A sketching algorithm is also introduced to further improve efficiency when using the $\ell_2$ loss function. To demonstrate the efficacy of our methods, we provide illustrative examples using both synthetic data and an early childhood human microbiome dataset.
Authors: Yingzhen Yang, Ping Li
Abstract: We study nonparametric regression by an over-parameterized two-layer neural network trained by gradient descent (GD) in this paper. We show that, if the neural network is trained by GD with early stopping, then the trained network renders a sharp rate of the nonparametric regression risk of $\mathcal{O}(\epsilon_n^2)$, which is the same rate as that for the classical kernel regression trained by GD with early stopping, where $\epsilon_n$ is the critical population rate of the Neural Tangent Kernel (NTK) associated with the network and $n$ is the size of the training data. It is remarked that our result does not require distributional assumptions about the covariate as long as the covariate is bounded, in a strong contrast with many existing results which rely on specific distributions of the covariates such as the spherical uniform data distribution or distributions satisfying certain restrictive conditions. The rate $\mathcal{O}(\epsilon_n^2)$ is known to be minimax optimal for specific cases, such as the case that the NTK has a polynomial eigenvalue decay rate which happens under certain distributional assumptions on the covariates. Our result formally fills the gap between training a classical kernel regression model and training an over-parameterized but finite-width neural network by GD for nonparametric regression without distributional assumptions on the bounded covariate. We also provide confirmative answers to certain open questions or address particular concerns in the literature of training over-parameterized neural networks by GD with early stopping for nonparametric regression, including the characterization of the stopping time, the lower bound for the network width, and the constant learning rate used in GD.
Authors: Zhihao Zhu, Yi Yang, Defu Lian
Abstract: Training Data Detection (TDD) is a task aimed at determining whether a specific data instance is used to train a machine learning model. In the computer security literature, TDD is also referred to as Membership Inference Attack (MIA). Given its potential to assess the risks of training data breaches, ensure copyright authentication, and verify model unlearning, TDD has garnered significant attention in recent years, leading to the development of numerous methods. Despite these advancements, there is no comprehensive benchmark to thoroughly evaluate the effectiveness of TDD methods. In this work, we introduce TDDBench, which consists of 13 datasets spanning three data modalities: image, tabular, and text. We benchmark 21 different TDD methods across four detection paradigms and evaluate their performance from five perspectives: average detection performance, best detection performance, memory consumption, and computational efficiency in both time and memory. With TDDBench, researchers can identify bottlenecks and areas for improvement in TDD algorithms, while practitioners can make informed trade-offs between effectiveness and efficiency when selecting TDD algorithms for specific use cases. Our extensive experiments also reveal the generally unsatisfactory performance of TDD algorithms across different datasets. To enhance accessibility and reproducibility, we open-source TDDBench for the research community at https://github.com/zzh9568/TDDBench.
Authors: David M. Bossens, Kishor Bharti, Jayne Thompson
Abstract: Parametrised quantum circuits offer expressive and data-efficient representations for machine learning. Due to quantum states residing in a high-dimensional Hilbert space, parametrised quantum circuits have a natural interpretation in terms of kernel methods. The representation of quantum circuits in terms of quantum kernels has been studied widely in quantum supervised learning, but has been overlooked in the context of quantum RL. This paper proposes the use of kernel policies and quantum policy gradient algorithms for quantum-accessible environments. After discussing the properties of such policies and a demonstration of classical policy gradient on a coherent policy in a quantum environment, we propose parametric and non-parametric policy gradient and actor-critic algorithms with quantum kernel policies in quantum environments. This approach, implemented with both numerical and analytical quantum policy gradient techniques, allows exploiting the many advantages of kernel methods, including data-driven forms for functions (and their gradients) as well as tunable expressiveness. The proposed approach is suitable for vector-valued action spaces and each of the formulations demonstrates a quadratic reduction in query complexity compared to their classical counterparts. We propose actor-critic algorithms based on stochastic policy gradient, deterministic policy gradient, and natural policy gradient, and demonstrate additional query complexity reductions compared to quantum policy gradient algorithms under favourable conditions.
Authors: Elie Azeraf
Abstract: The Pairwise Markov Chain (PMC) is a probabilistic graphical model extending the well-known Hidden Markov Model. This model, although highly effective for many tasks, has been scarcely utilized for continuous value prediction. This is mainly due to the issue of modeling observations inherent in generative probabilistic models. In this paper, we introduce a new algorithm for prediction with the PMC. On the one hand, this algorithm allows circumventing the feature problem, thus fully exploiting the capabilities of the PMC. On the other hand, it enables the PMC to extend any predictive model by introducing hidden states, updated at each time step, and allowing the introduction of non-stationarity for any model. We apply the PMC with its new algorithm for volatility forecasting, which we compare to the highly popular GARCH(1,1) and feedforward neural models across numerous pairs. This is particularly relevant given the regime changes that we can observe in volatility. For each scenario, our algorithm enhances the performance of the extended model, demonstrating the value of our approach.
Authors: Haoyang Li, Marko Stamenkovic, Alexander Shmakov, Michael Fenton, Darius Shih-Chieh Chao, Kaitlyn Maiya White, Caden Mikkelsen, Jovan Mitic, Cristina Mantilla Suarez, Melissa Quinnan, Greg Landsberg, Harvey Newman, Pierre Baldi, Daniel Whiteson, Javier Duarte
Abstract: The production of multiple Higgs bosons at the CERN LHC provides a direct way to measure the trilinear and quartic Higgs self-interaction strengths as well as potential access to beyond the standard model effects that can enhance production at large transverse momentum $p_{\mathrm{T}}$. The largest event fraction arises from the fully hadronic final state in which every Higgs boson decays to a bottom quark-antiquark pair ($b\bar{b}$). This introduces a combinatorial challenge known as the \emph{jet assignment problem}: assigning jets to sets representing Higgs boson candidates. Symmetry-preserving attention networks (SPA-Nets) have been been developed to address this challenge. However, the complexity of jet assignment increases when simultaneously considering both $H\rightarrow b\bar{b}$ reconstruction possibilities, i.e., two "resolved" small-radius jets each containing a shower initiated by a $b$-quark or one "boosted" large-radius jet containing a merged shower initiated by a $b\bar{b}$ pair. The latter improves the reconstruction efficiency at high $p_{\mathrm{T}}$. In this work, we introduce a generalization to the SPA-Net approach to simultaneously consider both boosted and resolved reconstruction possibilities and unambiguously interpret an event as "fully resolved'', "fully boosted", or in between. We report the performance of baseline methods, the original SPA-Net approach, and our generalized version on nonresonant $HH$ and $HHH$ production at the LHC. Considering both boosted and resolved topologies, our SPA-Net approach increases the Higgs boson reconstruction purity by 57--62\% and the efficiency by 23--38\% compared to the baseline method depending on the final state.
Authors: Donald Shenaj, Ondrej Bohdal, Mete Ozay, Pietro Zanuttigh, Umberto Michieli
Abstract: Recent advancements in image generation models have enabled personalized image creation with both user-defined subjects (content) and styles. Prior works achieved personalization by merging corresponding low-rank adapters (LoRAs) through optimization-based methods, which are computationally demanding and unsuitable for real-time use on resource-constrained devices like smartphones. To address this, we introduce LoRA$.$rar, a method that not only improves image quality but also achieves a remarkable speedup of over $4000\times$ in the merging process. We collect a dataset of style and subject LoRAs and pre-train a hypernetwork on a diverse set of content-style LoRA pairs, learning an efficient merging strategy that generalizes to new, unseen content-style pairs, enabling fast, high-quality personalization. Moreover, we identify limitations in existing evaluation metrics for content-style quality and propose a new protocol using multimodal large language models (MLLMs) for more accurate assessment. Our method significantly outperforms the current state of the art in both content and style fidelity, as validated by MLLM assessments and human evaluations.
Authors: Andrew H. Zhang, Alex He-Mo, Richard Fei Yin, Chunlin Li, Yuzhi Tang, Dharmendra Gurve, Veronique van der Horst, Aron S. Buchman, Nasim Montazeri Ghahjaverestan, Maged Goubran, Bo Wang, Andrew S. P. Lim
Abstract: Study Objectives: We investigate a Mamba-based deep learning approach for sleep staging on signals from ANNE One (Sibel Health, Evanston, IL), a non-intrusive dual-module wireless wearable system measuring chest electrocardiography (ECG), triaxial accelerometry, and chest temperature, and finger photoplethysmography and finger temperature. Methods: We obtained wearable sensor recordings from 357 adults undergoing concurrent polysomnography (PSG) at a tertiary care sleep lab. Each PSG recording was manually scored and these annotations served as ground truth labels for training and evaluation of our models. PSG and wearable sensor data were automatically aligned using their ECG channels with manual confirmation by visual inspection. We trained a Mamba-based recurrent neural network architecture on these recordings. Ensembling of model variants with similar architectures was performed. Results: After ensembling, the model attains a 3-class (wake, non rapid eye movement [NREM] sleep, rapid eye movement [REM] sleep) balanced accuracy of 84.02%, F1 score of 84.23%, Cohen's $\kappa$ of 72.89%, and a Matthews correlation coefficient (MCC) score of 73.00%; a 4-class (wake, light NREM [N1/N2], deep NREM [N3], REM) balanced accuracy of 75.30%, F1 score of 74.10%, Cohen's $\kappa$ of 61.51%, and MCC score of 61.95%; a 5-class (wake, N1, N2, N3, REM) balanced accuracy of 65.11%, F1 score of 66.15%, Cohen's $\kappa$ of 53.23%, MCC score of 54.38%. Conclusions: Our Mamba-based deep learning model can successfully infer major sleep stages from the ANNE One, a wearable system without electroencephalography (EEG), and can be applied to data from adults attending a tertiary care sleep clinic.
Authors: Yilie Huang, Yanwei Jia, Xun Yu Zhou
Abstract: We study continuous-time mean--variance portfolio selection in markets where stock prices are diffusion processes driven by observable factors that are also diffusion processes, yet the coefficients of these processes are unknown. Based on the recently developed reinforcement learning (RL) theory for diffusion processes, we present a general data-driven RL algorithm that learns the pre-committed investment strategy directly without attempting to learn or estimate the market coefficients. For multi-stock Black--Scholes markets without factors, we further devise a baseline algorithm and prove its performance guarantee by deriving a sublinear regret bound in terms of the Sharpe ratio. For performance enhancement and practical implementation, we modify the baseline algorithm and carry out an extensive empirical study to compare its performance, in terms of a host of common metrics, with a large number of widely employed portfolio allocation strategies on S\&P 500 constituents. The results demonstrate that the proposed continuous-time RL strategy is consistently among the best, especially in a volatile bear market, and decisively outperforms the model-based continuous-time counterparts by significant margins.
Authors: Scott Emmons, Caspar Oesterheld, Vincent Conitzer, Stuart Russell
Abstract: We study partially observable assistance games (POAGs), a model of the human-AI value alignment problem which allows the human and the AI assistant to have partial observations. Motivated by concerns of AI deception, we study a qualitatively new phenomenon made possible by partial observability: would an AI assistant ever have an incentive to interfere with the human's observations? First, we prove that sometimes an optimal assistant must take observation-interfering actions, even when the human is playing optimally, and even when there are otherwise-equivalent actions available that do not interfere with observations. Though this result seems to contradict the classic theorem from single-agent decision making that the value of information is nonnegative, we resolve this seeming contradiction by developing a notion of interference defined on entire policies. This can be viewed as an extension of the classic result that the value of information is nonnegative into the cooperative multiagent setting. Second, we prove that if the human is simply making decisions based on their immediate outcomes, the assistant might need to interfere with observations as a way to query the human's preferences. We show that this incentive for interference goes away if the human is playing optimally, or if we introduce a communication channel for the human to communicate their preferences to the assistant. Third, we show that if the human acts according to the Boltzmann model of irrationality, this can create an incentive for the assistant to interfere with observations. Finally, we use an experimental model to analyze tradeoffs faced by the AI assistant in practice when considering whether or not to take observation-interfering actions.
Authors: Hao Xie, Saburo Howard, Guglielmo Mazzola
Abstract: Accurate determination of the equation of state of dense hydrogen is essential for understanding gas giants. Currently, there is still no consensus on methods for calculating its entropy, which play a fundamental role and can result in qualitatively different predictions for Jupiter's interior. Here, we investigate various aspects of entropy calculation for dense hydrogen based on ab initio molecular dynamics simulations. Specifically, we employ the recently developed flow matching method to validate the accuracy of the traditional thermodynamic integration approach. We then clearly identify pitfalls in previous attempts and propose a reliable framework for constructing the hydrogen equation of state, which is accurate and thermodynamically consistent across a wide range of temperature and pressure conditions. This allows us to conclusively address the long-standing discrepancies in Jupiter's adiabat among earlier studies, demonstrating the potential of our approach for providing reliable equations of state of diverse materials.
Authors: Bo Gao, Michael W. Spratling
Abstract: Large language models have achieved remarkable success in recent years, primarily due to the implementation of self-attention mechanisms. However, traditional Softmax attention suffers from numerical instability and reduced performance as the length of inference tokens increases. This paper addresses these issues by proposing a new design principle for attention, viewing it as a two-stage process. We first decompose the Softmax operation into a non-linear positivity transformation and an $l_1$-normalisation step, identifying the latter as essential for maintaining model performance. In the first stage, we replace the standard exponential function with the more numerically stable Softplus activation and introduce a dynamic scale factor based on invariance entropy, creating a novel attention mechanism that outperforms conventional Softmax attention. In the second stage, we introduce a re-weighting mechanism that sharpens the attention distribution, amplifying significant weights while diminishing weaker ones. This enables the model to concentrate more effectively on relevant tokens and fundamentally improves length extrapolation. When combined, this two-stage approach ensures numerical stability and dramatically improves length extrapolation, maintaining a nearly constant validation loss at 16$\times$ the training length while achieving superior results on challenging long-context retrieval tasks and standard downstream benchmarks.
Authors: Shuo Shao, Haozhe Zhu, Yiming Li, Hongwei Yao, Tianwei Zhang, Zhan Qin
Abstract: Model fingerprinting is a widely adopted approach to safeguard the intellectual property rights of open-source models by preventing their unauthorized reuse. It is promising and convenient since it does not necessitate modifying the protected model. In this paper, we revisit existing fingerprinting methods and reveal that they are vulnerable to false claim attacks where adversaries falsely assert ownership of any third-party model. We demonstrate that this vulnerability mostly stems from their untargeted nature, where they generally compare the outputs of given samples on different models instead of the similarities to specific references. Motivated by these findings, we propose a targeted fingerprinting paradigm (i.e., FIT-Print) to counteract false claim attacks. Specifically, FIT-Print transforms the fingerprint into a targeted signature via optimization. Building on the principles of FIT-Print, we develop bit-wise and list-wise black-box model fingerprinting methods, i.e., FIT-ModelDiff and FIT-LIME, which exploit the distance between model outputs and the feature attribution of specific samples as the fingerprint, respectively. Extensive experiments on benchmark models and datasets verify the effectiveness, conferrability, and resistance to false claim attacks of our FIT-Print.
Authors: Michael Birsak, John Femiani, Biao Zhang, Peter Wonka
Abstract: Assigning realistic materials to 3D models remains a significant challenge in computer graphics. We propose MatCLIP, a novel method that extracts shape- and lighting-insensitive descriptors of Physically Based Rendering (PBR) materials to assign plausible textures to 3D objects based on images, such as the output of Latent Diffusion Models (LDMs) or photographs. Matching PBR materials to static images is challenging because the PBR representation captures the dynamic appearance of materials under varying viewing angles, shapes, and lighting conditions. By extending an Alpha-CLIP-based model on material renderings across diverse shapes and lighting, and encoding multiple viewing conditions for PBR materials, our approach generates descriptors that bridge the domains of PBR representations with photographs or renderings, including LDM outputs. This enables consistent material assignments without requiring explicit knowledge of material relationships between different parts of an object. MatCLIP achieves a top-1 classification accuracy of 76.6%, outperforming state-of-the-art methods such as PhotoShape and MatAtlas by over 15 percentage points on publicly available datasets. Our method can be used to construct material assignments for 3D shape datasets such as ShapeNet, 3DCoMPaT++, and Objaverse. All code and data will be released.
Authors: Rui Min, Tianyu Pang, Chao Du, Qian Liu, Minhao Cheng, Min Lin
Abstract: Chatbot Arena is a popular platform for evaluating LLMs by pairwise battles, where users vote for their preferred response from two randomly sampled anonymous models. While Chatbot Arena is widely regarded as a reliable LLM ranking leaderboard, we show that crowdsourced voting can be rigged to improve (or decrease) the ranking of a target model $m_{t}$. We first introduce a straightforward target-only rigging strategy that focuses on new battles involving $m_{t}$, identifying it via watermarking or a binary classifier, and exclusively voting for $m_{t}$ wins. However, this strategy is practically inefficient because there are over $190$ models on Chatbot Arena and on average only about $1\%$ of new battles will involve $m_{t}$. To overcome this, we propose omnipresent rigging strategies, exploiting the Elo rating mechanism of Chatbot Arena that any new vote on a battle can influence the ranking of the target model $m_{t}$, even if $m_{t}$ is not directly involved in the battle. We conduct experiments on around $1.7$ million historical votes from the Chatbot Arena Notebook, showing that omnipresent rigging strategies can improve model rankings by rigging only hundreds of new votes. While we have evaluated several defense mechanisms, our findings highlight the importance of continued efforts to prevent vote rigging. Our code is available at https://github.com/sail-sg/Rigging-ChatbotArena.
Authors: Xuefeng Liu, Songhao Jiang, Ian Foster, Jinbo Xu, Rick Stevens
Abstract: Drug optimization has become increasingly crucial in light of fast-mutating virus strains and drug-resistant cancer cells. Nevertheless, it remains challenging as it necessitates retaining the beneficial properties of the original drug while simultaneously enhancing desired attributes beyond its scope. In this work, we aim to tackle this challenge by introducing ScaffoldGPT, a novel Generative Pretrained Transformer (GPT) designed for drug optimization based on molecular scaffolds. Our work comprises three key components: (1) A three-stage drug optimization approach that integrates pretraining, finetuning, and decoding optimization. (2) A novel two-phase incremental pre-training strategy for scaffold-based drug optimization. (3) A token-level decoding optimization strategy, Top-N, that enabling controlled, reward-guided generation using the pretrained or finetuned GPT. We demonstrate via a comprehensive evaluation on COVID and cancer benchmarks that ScaffoldGPT outperforms the competing baselines in drug optimization benchmarks, while excelling in preserving original functional scaffold and enhancing desired properties.
Authors: Yuan Sui, Yufei He, Tri Cao, Simeng Han, Yulin Chen, Bryan Hooi
Abstract: Large Language Models (LLMs) struggle with high computational time and error propagation during inference time, especially for complex tasks like math, puzzles, or coding requiring multi-step thinking. While existing reasoning models with chain-of-thoughts (CoT) can enable LLMs to do step-wise analysis and reflection, they often face the issue of wasting computation on less productive solutions and fail to make progress during inference time. In this paper, we propose Meta-Reasoner, a new framework to enable LLMs ``Think about how to think'', i.e., optimize the inference compute by adjusting strategies on how to reason during inference time. Inspired by dual-process theory, our method decouples the high-level strategy generation (e.g., backtracking, switching approaches, or restarting) from stepwise CoT generation via a lightweight progress report. The strategy module only consider the summarized version from the previous CoTs to propose new strategies accordingly. We employ the contextual multi-armed bandits (CMABs) for this module to iteratively evaluate the previous reasoning states and dynamically adjust the strategy to avoid reasoning get stuck in less productive paths during inference. Evaluations on math problems (e.g., Game-of-24, TheoremQA) and scientific problems (e.g., SciBench) demonstrate that our method improves performance by 9-12\% over previous SOTA methods while reducing inference time by 28-35\%. This approach also generalizes to other domains like creative writing, demonstrating its versatility for diverse reasoning-intensive problems using LLMs.
Authors: Ojonugwa Oluwafemi Ejiga Peter, Md Mahmudur Rahman, Fahmi Khalifa
Abstract: The MEDVQA-GI challenge addresses the integration of AI-driven text-to-image generative models in medical diagnostics, aiming to enhance diagnostic capabilities through synthetic image generation. Existing methods primarily focus on static image analysis and lack the dynamic generation of medical imagery from textual descriptions. This study intends to partially close this gap by introducing a novel approach based on fine-tuned generative models to generate dynamic, scalable, and precise images from textual descriptions. Particularly, our system integrates fine-tuned Stable Diffusion and DreamBooth models, as well as Low-Rank Adaptation (LORA), to generate high-fidelity medical images. The problem is around two sub-tasks namely: image synthesis (IS) and optimal prompt production (OPG). The former creates medical images via verbal prompts, whereas the latter provides prompts that produce high-quality images in specified categories. The study emphasizes the limitations of traditional medical image generation methods, such as hand sketching, constrained datasets, static procedures, and generic models. Our evaluation measures showed that Stable Diffusion surpasses CLIP and DreamBooth + LORA in terms of producing high-quality, diversified images. Specifically, Stable Diffusion had the lowest Fr\'echet Inception Distance (FID) scores (0.099 for single center, 0.064 for multi-center, and 0.067 for combined), indicating higher image quality. Furthermore, it had the highest average Inception Score (2.327 across all datasets), indicating exceptional diversity and quality. This advances the field of AI-powered medical diagnosis. Future research will concentrate on model refining, dataset augmentation, and ethical considerations for efficiently implementing these advances into clinical practice
Authors: Antonin Schrab
Abstract: This article provides a practical introduction to kernel discrepancies, focusing on the Maximum Mean Discrepancy (MMD), the Hilbert-Schmidt Independence Criterion (HSIC), and the Kernel Stein Discrepancy (KSD). Various estimators for these discrepancies are presented, including the commonly-used V-statistics and U-statistics, as well as several forms of the more computationally-efficient incomplete U-statistics. The importance of the choice of kernel bandwidth is stressed, showing how it affects the behaviour of the discrepancy estimation. Adaptive estimators are introduced, which combine multiple estimators with various kernels, addressing the problem of kernel selection.
Authors: Nirmit Joshi, Gal Vardi, Adam Block, Surbhi Goel, Zhiyuan Li, Theodor Misiakiewicz, Nathan Srebro
Abstract: For a given base class of sequence-to-next-token generators, we consider learning prompt-to-answer mappings obtained by iterating a fixed, time-invariant generator for multiple steps, thus generating a chain-of-thought, and then taking the final token as the answer. We formalize the learning problems both when the chain-of-thought is observed and when training only on prompt-answer pairs, with the chain-of-thought latent. We analyze the sample and computational complexity both in terms of general properties of the base class (e.g. its VC dimension) and for specific base classes such as linear thresholds. We present a simple base class that allows for universal representability and computationally tractable chain-of-thought learning. Central to our development is that time invariance allows for sample complexity that is independent of the length of the chain-of-thought. Attention arises naturally in our construction.
Authors: Bikram Das, Rupchand Sutradhar, D C Dalal
Abstract: Accurate forecasting of viral disease outbreaks is crucial for guiding public health responses and preventing widespread loss of life. In recent years, Physics-Informed Neural Networks (PINNs) have emerged as a promising framework that can capture the intricate dynamics of viral infection and reliably predict its future progression. However, despite notable advances, the application of PINNs in disease modeling remains limited. Standard PINNs are effective in simulating disease dynamics through forward modeling but often face challenges in estimating key biological parameters from sparse or noisy experimental data when applied in an inverse framework. To overcome these limitations, a recent extension known as Disease Informed Neural Networks (DINNs) has emerged, offering a more robust approach to parameter estimation tasks. In this work, we apply this DINNs technique on a recently proposed hepatitis B virus (HBV) infection dynamics model to predict infection transmission within the liver. This model consists of four compartments: uninfected and infected hepatocytes, rcDNA-containing capsids, and free viruses. Leveraging the power of DINNs, we study the impacts of (i) variations in parameter range, (ii) experimental noise in data, (iii) sample sizes, (iv) network architecture and (v) learning rate. We employ this methodology in experimental data collected from nine HBV-infected chimpanzees and observe that it reliably estimates the model parameters. DINNs can capture infection dynamics and predict their future progression even when data of some compartments of the system are missing. Additionally, it identifies the influential model parameters that determine whether the HBV infection is cleared or persists within the host.
Authors: Philipp Becker, Abhinav Mehrotra, Ruchika Chavhan, Malcolm Chadwick, Luca Morreale, Mehdi Noroozi, Alberto Gil Ramos, Sourav Bhattacharya
Abstract: Diffusion Transformers (DiTs) have emerged as a leading architecture for text-to-image synthesis, producing high-quality and photorealistic images. However, the quadratic scaling properties of the attention in DiTs hinder image generation with higher resolution or on devices with limited resources. This work introduces an efficient diffusion transformer (EDiT) to alleviate these efficiency bottlenecks in conventional DiTs and Multimodal DiTs (MM-DiTs). First, we present a novel linear compressed attention method that uses a multi-layer convolutional network to modulate queries with local information while keys and values are aggregated spatially. Second, we formulate a hybrid attention scheme for multimodal inputs that combines linear attention for image-to-image interactions and standard scaled dot-product attention for interactions involving prompts. Merging these two approaches leads to an expressive, linear-time Multimodal Efficient Diffusion Transformer (MM-EDiT). We demonstrate the effectiveness of the EDiT and MM-EDiT architectures by integrating them into PixArt-Sigma (conventional DiT) and Stable Diffusion 3.5-Medium (MM-DiT), achieving up to 2.2x speedup with comparable image quality after distillation.
Authors: Liangzhi Shi, Yulin Liu, Lingqi Zeng, Bo Ai, Zhengdong Hong, Hao Su
Abstract: How can robots learn dexterous grasping skills efficiently and apply them adaptively based on user instructions? This work tackles two key challenges: efficient skill acquisition from limited human demonstrations and context-driven skill selection. We introduce AdaDexGrasp, a framework that learns a library of grasping skills from a single human demonstration per skill and selects the most suitable one using a vision-language model (VLM). To improve sample efficiency, we propose a trajectory following reward that guides reinforcement learning (RL) toward states close to a human demonstration while allowing flexibility in exploration. To learn beyond the single demonstration, we employ curriculum learning, progressively increasing object pose variations to enhance robustness. At deployment, a VLM retrieves the appropriate skill based on user instructions, bridging low-level learned skills with high-level intent. We evaluate AdaDexGrasp in both simulation and real-world settings, showing that our approach significantly improves RL efficiency and enables learning human-like grasp strategies across varied object configurations. Finally, we demonstrate zero-shot transfer of our learned policies to a real-world PSYONIC Ability Hand, with a 90% success rate across objects, significantly outperforming the baseline.
Authors: Mikel Zhobro, Andreas Ren\'e Geist, Georg Martius
Abstract: Realistic simulation is critical for applications ranging from robotics to animation. Learned simulators have emerged as a possibility to capture real world physics directly from video data, but very often require privileged information such as depth information, particle tracks and hand-engineered features to maintain spatial and temporal consistency. These strong inductive biases or ground truth 3D information help in domains where data is sparse but limit scalability and generalization in data rich regimes. To overcome the key limitations, we propose 3DGSim, a learned 3D simulator that directly learns physical interactions from multi-view RGB videos. 3DGSim unifies 3D scene reconstruction, particle dynamics prediction and video synthesis into an end-to-end trained framework. It adopts MVSplat to learn a latent particle-based representation of 3D scenes, a Point Transformer for particle dynamics, a Temporal Merging module for consistent temporal aggregation and Gaussian Splatting to produce novel view renderings. By jointly training inverse rendering and dynamics forecasting, 3DGSim embeds the physical properties into point-wise latent features. This enables the model to capture diverse physical behaviors, from rigid to elastic, cloth-like dynamics, and boundary conditions (e.g. fixed cloth corner), along with realistic lighting effects that also generalize to unseen multibody interactions and novel scene edits.
Authors: Hongzhe Du, Weikai Li, Min Cai, Karim Saraipour, Zimin Zhang, Himabindu Lakkaraju, Yizhou Sun, Shichang Zhang
Abstract: Post-training is essential for the success of large language models (LLMs), transforming pre-trained base models into more useful and aligned post-trained models. While plenty of works have studied post-training algorithms and evaluated post-training models by their outputs, it remains understudied how post-training reshapes LLMs internally. In this paper, we compare base and post-trained LLMs mechanistically from four perspectives to better understand post-training effects. Our findings across model families and datasets reveal that: (1) Post-training does not change the factual knowledge storage locations, and it adapts knowledge representations from the base model while developing new knowledge representations; (2) Both truthfulness and refusal can be represented by vectors in the hidden representation space. The truthfulness direction is highly similar between the base and post-trained model, and it is effectively transferable for interventions; (3) The refusal direction is different between the base and post-trained models, and it shows limited forward transferability; (4) Differences in confidence between the base and post-trained models cannot be attributed to entropy neurons. Our study provides insights into the fundamental mechanisms preserved and altered during post-training, facilitates downstream tasks like model steering, and could potentially benefit future research in interpretability and LLM post-training. Our code is publicly available at https://github.com/HZD01/post-training-mechanistic-analysis.
URLs: https://github.com/HZD01/post-training-mechanistic-analysis.
Authors: Patrick Fernandes, Sweta Agrawal, Emmanouil Zaranis, Andr\'e F. T. Martins, Graham Neubig
Abstract: Despite the steady progress in machine translation evaluation, existing automatic metrics struggle to capture how well meaning is preserved beyond sentence boundaries. We posit that reliance on a single intrinsic quality score, trained to mimic human judgments, might be insufficient for evaluating translations of long, complex passages, and a more ``pragmatic'' approach that assesses how accurately key information is conveyed by a translation in context is needed. We introduce TREQA (Translation Evaluation via Question-Answering), a framework that extrinsically evaluates translation quality by assessing how accurately candidate translations answer reading comprehension questions that target key information in the original source or reference texts. In challenging domains that require long-range understanding, such as literary texts, we show that TREQA is competitive with and, in some cases, outperforms state-of-the-art neural and LLM-based metrics in ranking alternative paragraph-level translations, despite never being explicitly optimized to correlate with human judgments. Furthermore, the generated questions and answers offer interpretability: empirical analysis shows that they effectively target translation errors identified by experts in evaluated datasets. Our code is available at https://github.com/deep-spin/treqa
Authors: Atharva Nijasure, Tanya Chowdhury, James Allan
Abstract: We conduct a behavioral exploration of LoRA fine-tuned LLMs for Passage Reranking to understand how relevance signals are learned and deployed by Large Language Models. By fine-tuning Mistral-7B, LLaMA3.1-8B, and Pythia-6.9B on MS MARCO under diverse LoRA configurations, we investigate how relevance modeling evolves across checkpoints, the impact of LoRA rank (1, 2, 8, 32), and the relative importance of updated MHA vs. MLP components. Our ablations reveal which layers and projections within LoRA transformations are most critical for reranking accuracy. These findings offer fresh explanations into LoRA's adaptation mechanisms, setting the stage for deeper mechanistic studies in Information Retrieval. All models used in this study have been shared.
Authors: Md Rakibul Hasan, Md Zakir Hossain, Aneesh Krishna, Shafin Rahman, Tom Gedeon
Abstract: Detecting empathy from video interactions is an emerging area of research, particularly in healthcare and social robotics. However, privacy and ethical concerns often prevent the release of raw video data, with many datasets instead shared as pre-extracted tabular features. Previous work on such datasets has established classical tree-based models as the state of the art. Motivated by recent successes of large-scale foundation models for text, we investigate the potential of tabular foundation models (TFMs) for empathy detection from video-derived tabular data. Our proposed system, TFMPathy, is demonstrated with two recent TFMs (TabPFN v2 and TabICL) under both in-context learning and fine-tuning paradigms. On a public human-robot interaction benchmark, TFMPathy significantly improves empathy detection accuracy reported in the literature. While the established evaluation protocol in the literature does not ensure cross-subject generalisation, our evaluation scheme also captures such generalisation. We show that TFMPathy under a fine-tuning setup has better cross-subject generalisation capacity over baseline methods (accuracy: $0.590 \rightarrow 0.730$; AUC: $0.564 \rightarrow 0.669$). Given the ongoing privacy and ethical constraints around raw video sharing, the proposed TFMPathy system provides a practical and scalable path toward building AI systems dependent on human-centred video datasets. Our code is publicly available at https://github.com/hasan-rakibul/TFMPathy (will be made available upon acceptance of this paper).
Authors: Linjuan Fan, Di Wen, Kunyu Peng, Kailun Yang, Jiaming Zhang, Ruiping Liu, Yufan Chen, Junwei Zheng, Jiamin Wu, Xudong Han, Rainer Stiefelhagen
Abstract: As an open research topic in the field of deep learning, learning with noisy labels has attracted much attention and grown rapidly over the past ten years. Learning with label noise is crucial for driver distraction behavior recognition, as real-world video data often contains mislabeled samples, impacting model reliability and performance. However, label noise learning is barely explored in the driver activity recognition field. In this paper, we propose the first label noise learning approach for the driver activity recognition task. Based on the cluster assumption, we initially enable the model to learn clustering-friendly low-dimensional representations from given videos and assign the resultant embeddings into clusters. We subsequently perform co-refinement within each cluster to smooth the classifier outputs. Furthermore, we propose a flexible sample selection strategy that combines two selection criteria without relying on any hyperparameters to filter clean samples from the training dataset. We also incorporate a self-adaptive parameter into the sample selection process to enforce balancing across classes. A comprehensive variety of experiments on the public Drive&Act dataset for all granularity levels demonstrates the superior performance of our method in comparison with other label-denoising methods derived from the image classification field. The source code is available at https://github.com/ilonafan/DAR-noisy-labels.
Authors: Yunheng Zou, Austin H. Cheng, Abdulrahman Aldossary, Jiaru Bai, Shi Xuan Leong, Jorge Arturo Campos-Gonzalez-Angulo, Changhyeok Choi, Cher Tian Ser, Gary Tom, Andrew Wang, Zijian Zhang, Ilya Yakavets, Han Hao, Chris Crebolder, Varinia Bernales, Al\'an Aspuru-Guzik
Abstract: Computational chemistry tools are widely used to study the behaviour of chemical phenomena. Yet, the complexity of these tools can make them inaccessible to non-specialists and challenging even for experts. In this work, we introduce El Agente Q, an LLM-based multi-agent system that dynamically generates and executes quantum chemistry workflows from natural language user prompts. The system is built on a novel cognitive architecture featuring a hierarchical memory framework that enables flexible task decomposition, adaptive tool selection, post-analysis, and autonomous file handling and submission. El Agente Q is benchmarked on six university-level course exercises and two case studies, demonstrating robust problem-solving performance (averaging >87% task success) and adaptive error handling through in situ debugging. It also supports longer-term, multi-step task execution for more complex workflows, while maintaining transparency through detailed action trace logs. Together, these capabilities lay the foundation for increasingly autonomous and accessible quantum chemistry.
Authors: Qize Jiang, Linsey Pang, Alice Gatti, Mahima Aggarwal, Giovanna Vantini, Xiaosong Ma, Weiwei Sun, Sourav Medya, Sanjay Chawla
Abstract: Reinforcement Learning (RL) has proven to be a powerful tool for combinatorial optimization (CO) problems due to its ability to learn heuristics that can generalize across problem instances. However, integrating knowledge that will steer the RL framework for CO solutions towards domain appropriate outcomes remains a challenging task. In this paper, we propose RIDGECUT, the first RL framework that constrains the action space to enforce structure-aware partitioning in the Normalized Cut problem. Using transportation networks as a motivating example, we introduce a novel concept that leverages domain knowledge about urban road topology -- where natural partitions often take the form of concentric rings and radial wedges. Our method reshapes the graph into a linear or circular structure to simplify the partitioning task so that we can apply sequential transformers and enables efficient learning via Proximal Policy Optimization. The resulting partitions are not only aligned with expected spatial layouts but also achieve lower normalized cuts compared to existing methods. While we focus on traffic data, our approach is broadly applicable and offers a mechanism for embedding structural priors into RL for graph partitioning.
Authors: Kaveen Hiniduma, Zilinghan Li, Aditya Sinha, Ravi Madduri, Suren Byna
Abstract: Privacy-Preserving Federated Learning (PPFL) is a decentralized machine learning approach where multiple clients train a model collaboratively. PPFL preserves the privacy and security of a client's data without exchanging it. However, ensuring that data at each client is of high quality and ready for federated learning (FL) is a challenge due to restricted data access. In this paper, we introduce CADRE (Customizable Assurance of Data Readiness) for federated learning (FL), a novel framework that allows users to define custom data readiness (DR) metrics, rules, and remedies tailored to specific FL tasks. CADRE generates comprehensive DR reports based on the user-defined metrics, rules, and remedies to ensure datasets are prepared for FL while preserving privacy. We demonstrate a practical application of CADRE by integrating it into an existing PPFL framework. We conducted experiments across six datasets and addressed seven different DR issues. The results illustrate the versatility and effectiveness of CADRE in ensuring DR across various dimensions, including data quality, privacy, and fairness. This approach enhances the performance and reliability of FL models as well as utilizes valuable resources.
Authors: Edgar Welte, Rania Rayyes
Abstract: Dexterous manipulation is a crucial yet highly complex challenge in humanoid robotics, demanding precise, adaptable, and sample-efficient learning methods. As humanoid robots are usually designed to operate in human-centric environments and interact with everyday objects, mastering dexterous manipulation is critical for real-world deployment. Traditional approaches, such as reinforcement learning and imitation learning, have made significant strides, but they often struggle due to the unique challenges of real-world dexterous manipulation, including high-dimensional control, limited training data, and covariate shift. This survey provides a comprehensive overview of these challenges and reviews existing learning-based methods for real-world dexterous manipulation, spanning imitation learning, reinforcement learning, and hybrid approaches. A promising yet underexplored direction is interactive imitation learning, where human feedback actively refines a robots behavior during training. While interactive imitation learning has shown success in various robotic tasks, its application to dexterous manipulation remains limited. To address this gap, we examine current interactive imitation learning techniques applied to other robotic tasks and discuss how these methods can be adapted to enhance dexterous manipulation. By synthesizing state-of-the-art research, this paper highlights key challenges, identifies gaps in current methodologies, and outlines potential directions for leveraging interactive imitation learning to improve dexterous robotic skills.
Authors: Qihui Fan, Wenbo Li, Enfu Nan, Yixiao Chen, Lei Lu, Pu Zhao, Yanzhi Wang
Abstract: The growing popularity of social deduction games has created an increasing need for intelligent frameworks where humans can collaborate with AI agents, particularly in post-pandemic contexts with heightened psychological and social pressures. Social deduction games like Werewolf, traditionally played through verbal communication, present an ideal application for Large Language Models (LLMs) given their advanced reasoning and conversational capabilities. Prior studies have shown that LLMs can outperform humans in Werewolf games, but their reliance on external modules introduces latency that left their contribution in academic domain only, and omit such game should be user-facing. We propose \textbf{Verbal Werewolf}, a novel LLM-based Werewolf game system that optimizes two parallel pipelines: gameplay powered by state-of-the-art LLMs and a fine-tuned Text-to-Speech (TTS) module that brings text output to life. Our system operates in near real-time without external decision-making modules, leveraging the enhanced reasoning capabilities of modern LLMs like DeepSeek V3 to create a more engaging and anthropomorphic gaming experience that significantly improves user engagement compared to existing text-only frameworks.
Authors: Yongkang Xiao, Rui Zhang
Abstract: Multimodal knowledge graphs (MMKGs) enrich traditional knowledge graphs (KGs) by incorporating diverse modalities such as images and text. multimodal knowledge graph completion (MMKGC) seeks to exploit these heterogeneous signals to infer missing facts, thereby mitigating the intrinsic incompleteness of MMKGs. Existing MMKGC methods typically leverage only the information contained in the MMKGs under the closed-world assumption and adopt discriminative training objectives, which limits their reasoning capacity during completion. Recent large language models (LLMs), empowered by massive parameter scales and pretraining on vast corpora, have demonstrated strong reasoning abilities across various tasks. However, their potential in MMKGC remains largely unexplored. To bridge this gap, we propose HERGC, a flexible Heterogeneous Experts Representation and Generative Completion framework for MMKGs. HERGC first deploys a Heterogeneous Experts Representation Retriever that enriches and fuses multimodal information and retrieves a compact candidate set for each incomplete triple. It then uses a Generative LLM Predictor, implemented via either in-context learning or lightweight fine-tuning, to accurately identify the correct answer from these candidates. Extensive experiments on three standard MMKG benchmarks demonstrate HERGC's effectiveness and robustness, achieving superior performance over existing methods.
Authors: Emmanuel Hartman, Nicolas Charon
Abstract: Despite progress in the rapidly developing field of geometric deep learning, performing statistical analysis on geometric data--where each observation is a shape such as a curve, graph, or surface--remains challenging due to the non-Euclidean nature of shape spaces, which are defined as equivalence classes under invariance groups. Building machine learning frameworks that incorporate such invariances, notably to shape parametrization, is often crucial to ensure generalizability of the trained models to new observations. This work proposes \textit{SVarM} to exploit varifold representations of shapes as measures and their duality with test functions $h:\mathbb{R}^n \times S^{n-1} \rightarrow \mathbb{R}$. This method provides a general framework akin to linear support vector machines but operating instead over the infinite-dimensional space of varifolds. We develop classification and regression models on shape datasets by introducing a neural network-based representation of the trainable test function $h$. This approach demonstrates strong performance and robustness across various shape graph and surface datasets, achieving results comparable to state-of-the-art methods while significantly reducing the number of trainable parameters.
Authors: Payel Bhattacharjee, Fengwei Tian, Geoffrey D. Rubin, Joseph Y. Lo, Nirav Merchant, Heidi Hanson, John Gounley, Ravi Tandon
Abstract: Purpose: This study proposes a framework for fine-tuning large language models (LLMs) with differential privacy (DP) to perform multi-abnormality classification on radiology report text. By injecting calibrated noise during fine-tuning, the framework seeks to mitigate the privacy risks associated with sensitive patient data and protect against data leakage while maintaining classification performance. Materials and Methods: We used 50,232 radiology reports from the publicly available MIMIC-CXR chest radiography and CT-RATE computed tomography datasets, collected between 2011 and 2019. Fine-tuning of LLMs was conducted to classify 14 labels from MIMIC-CXR dataset, and 18 labels from CT-RATE dataset using Differentially Private Low-Rank Adaptation (DP-LoRA) in high and moderate privacy regimes (across a range of privacy budgets = {0.01, 0.1, 1.0, 10.0}). Model performance was evaluated using weighted F1 score across three model architectures: BERT-medium, BERT-small, and ALBERT-base. Statistical analyses compared model performance across different privacy levels to quantify the privacy-utility trade-off. Results: We observe a clear privacy-utility trade-off through our experiments on 2 different datasets and 3 different models. Under moderate privacy guarantees the DP fine-tuned models achieved comparable weighted F1 scores of 0.88 on MIMIC-CXR and 0.59 on CT-RATE, compared to non-private LoRA baselines of 0.90 and 0.78, respectively. Conclusion: Differentially private fine-tuning using LoRA enables effective and privacy-preserving multi-abnormality classification from radiology reports, addressing a key challenge in fine-tuning LLMs on sensitive medical data.
Authors: Renato Cordeiro Ferreira (Jheronimus Academy of Data Science, Technical University of Eindhoven, Tilburg University), Rowanne Trapmann (Jheronimus Academy of Data Science, Technical University of Eindhoven, Tilburg University), Willem-Jan van den Heuvel (Jheronimus Academy of Data Science, Technical University of Eindhoven, Tilburg University)
Abstract: This case study describes challenges and lessons learned on building Ocean Guard: a Machine Learning-Enabled System (MLES) for anomaly detection in the maritime domain. First, the paper presents the system's specification, and architecture. Ocean Guard was designed with a microservices' architecture to enable multiple teams to work on the project in parallel. Then, the paper discusses how the developers adapted contract-based design to MLOps for achieving that goal. As a MLES, Ocean Guard employs code, model, and data contracts to establish guidelines between its services. This case study hopes to inspire software engineers, machine learning engineers, and data scientists to leverage similar approaches for their systems.
Authors: Tengfei Lyu, Weijia Zhang, Hao Liu
Abstract: Subseasonal-to-seasonal (S2S) forecasting, which predicts climate conditions from several weeks to months in advance, represents a critical frontier for agricultural planning, energy management, and disaster preparedness. However, it remains one of the most challenging problems in atmospheric science, due to the chaotic dynamics of atmospheric systems and complex interactions across multiple scales. Current approaches often fail to explicitly model underlying physical processes and teleconnections that are crucial at S2S timescales. We introduce \textbf{TelePiT}, a novel deep learning architecture that enhances global S2S forecasting through integrated multi-scale physics and teleconnection awareness. Our approach consists of three key components: (1) Spherical Harmonic Embedding, which accurately encodes global atmospheric variables onto spherical geometry; (2) Multi-Scale Physics-Informed Neural ODE, which explicitly captures atmospheric physical processes across multiple learnable frequency bands; (3) Teleconnection-Aware Transformer, which models critical global climate interactions through explicitly modeling teleconnection patterns into the self-attention. Extensive experiments demonstrate that \textbf{TelePiT} significantly outperforms state-of-the-art data-driven baselines and operational numerical weather prediction systems across all forecast horizons, marking a significant advance toward reliable S2S forecasting.
Authors: Rui Zhang, Qi Meng, Han Wan, Yang Liu, Zhi-Ming Ma, Hao Sun
Abstract: Computational fluid dynamics (CFD) drives progress in numerous scientific and engineering fields, yet high-fidelity simulations remain computationally prohibitive. While machine learning approaches offer computing acceleration, they typically specialize in single physical systems or require extensive training data, hindering their applicability in highly nonlinear and 3D flow scenarios. To overcome these limitations, we propose OmniFluids, a pure physics pre-trained model that captures fundamental fluid dynamics laws and adapts efficiently to diverse downstream tasks with minimal data. We develop a training framework combining physics-only pre-training, coarse-grid operator distillation, and few-shot fine-tuning. This enables OmniFluids to retain broad physics knowledge while delivering fast and accurate predictions. Architecturally, OmniFluids integrates a mixture of operators, a multi-frame decoder, and factorized Fourier layers, seamlessly incorporating physics-based supervision while allowing efficient and scalable modeling of diverse tasks. Extensive tests on a broad range of 2D and 3D benchmarks show that OmniFluids outperforms state-of-the-art AI-driven methods in terms of flow field prediction and turbulence statistics. It delivers 10--100$\times$ speedups over traditional solvers while maintaining a comparable accuracy and accurately identifies unknown physical parameters from sparse, noisy data. This work demonstrates the potential of training a unified CFD solver exclusively from physics knowledge, offering a new approach for efficient and generalizable modeling across complex fluid systems.
Authors: Nuwan Bandara, Thivya Kandappu, Archan Misra
Abstract: Event-based eye tracking holds significant promise for fine-grained cognitive state inference, offering high temporal resolution and robustness to motion artifacts, critical features for decoding subtle mental states such as attention, confusion, or fatigue. In this work, we introduce a model-agnostic, inference-time refinement framework designed to enhance the output of existing event-based gaze estimation models without modifying their architecture or requiring retraining. Our method comprises two key post-processing modules: (i) Motion-Aware Median Filtering, which suppresses blink-induced spikes while preserving natural gaze dynamics, and (ii) Optical Flow-Based Local Refinement, which aligns gaze predictions with cumulative event motion to reduce spatial jitter and temporal discontinuities. To complement traditional spatial accuracy metrics, we propose a novel Jitter Metric that captures the temporal smoothness of predicted gaze trajectories based on velocity regularity and local signal complexity. Together, these contributions significantly improve the consistency of event-based gaze signals, making them better suited for downstream tasks such as micro-expression analysis and mind-state decoding. Our results demonstrate consistent improvements across multiple baseline models on controlled datasets, laying the groundwork for future integration with multimodal affect recognition systems in real-world environments. Our code implementations can be found at https://github.com/eye-tracking-for-physiological-sensing/EyeLoRiN.
URLs: https://github.com/eye-tracking-for-physiological-sensing/EyeLoRiN.
Authors: Andrew C. Cullen, Paul Montague, Sarah M. Erfani, Benjamin I. P. Rubinstein
Abstract: While certified robustness is widely promoted as a solution to adversarial examples in Artificial Intelligence systems, significant challenges remain before these techniques can be meaningfully deployed in real-world applications. We identify critical gaps in current research, including the paradox of detection without distinction, the lack of clear criteria for practitioners to evaluate certification schemes, and the potential security risks arising from users' expectations surrounding ``guaranteed" robustness claims. These create an alignment issue between how certifications are presented and perceived, relative to their actual capabilities. This position paper is a call to arms for the certification research community, proposing concrete steps to address these fundamental challenges and advance the field toward practical applicability.
Authors: Valentin Six, Evan Dufraisse, Ga\"el de Chalendar
Abstract: Large Language Models (LLMs) excel at many Natural Language Processing (NLP) tasks, but struggle with multi-hop reasoning and factual consistency, limiting their effectiveness on knowledge-intensive tasks like complex question answering (QA). Linking Knowledge Graphs (KG) and LLMs has shown promising results, but LLMs generally lack the ability to reason efficiently over graph-structured information. To address this challenge, we introduce DAGR, a retrieval method that leverages both complex questions and their decomposition in subquestions to extract relevant, linked textual subgraphs. DAGR first breaks down complex queries, retrieves subgraphs guided by a weighted similarity function over both the original and decomposed queries, and creates a question-specific knowledge graph to guide answer generation. The resulting Graph-RAG pipeline is suited to handle complex multi-hop questions and effectively reason over graph-structured data. We evaluate DAGR on standard multi-hop QA benchmarks and show that it achieves comparable or superior performance to competitive existing methods, using smaller models and fewer LLM calls.
Authors: Kenric P. Nelson
Abstract: The coupled entropy is proven to correct a flaw in the derivation of the Tsallis entropy and thereby solidify the theoretical foundations for analyzing the uncertainty of complex systems. The Tsallis entropy originated from considering power probabilities $p_i^q$ in which \textit{q} independent, identically-distributed random variables share the same state. The maximum entropy distribution was derived to be a \textit{q}-exponential, which is a member of the shape ($\kappa$), scale ($\sigma$) distributions. Unfortunately, the $q$-exponential parameters were treated as though valid substitutes for the shape and scale. This flaw causes a misinterpretation of the generalized temperature and an imprecise derivation of the generalized entropy. The coupled entropy is derived from the generalized Pareto distribution (GPD) and the Student's t distribution, whose shape derives from nonlinear sources and scale derives from linear sources of uncertainty. The Tsallis entropy of the GPD converges to one as $\kappa\rightarrow\infty$, which makes it too cold. The normalized Tsallis entropy (NTE) introduces a nonlinear term multiplying the scale and the coupling, making it too hot. The coupled entropy provides perfect balance, ranging from $\ln \sigma$ for $\kappa=0$ to $\sigma$ as $\kappa\rightarrow\infty$. One could say, the coupled entropy allows scientists, engineers, and analysts to eat their porridge, confident that its measure of uncertainty reflects the mathematical physics of the scale of non-exponential distributions while minimizing the dependence on the shape or nonlinear coupling. Examples of complex systems design including a coupled variation inference algorithm are reviewed.
Authors: Zhaoyang Xu, Yunbo Liu
Abstract: Identifying suitable machine learning paradigms for intrusion detection remains critical for building effective and generalizable security solutions. In this study, we present a controlled comparison of four representative models - Multi-Layer Perceptron (MLP), 1D Convolutional Neural Network (CNN), One-Class Support Vector Machine (OCSVM) and Local Outlier Factor (LOF) - on the CICIDS2017 dataset under two scenarios: detecting known attack types and generalizing to previously unseen threats. Our results show that supervised MLP and CNN achieve near-perfect accuracy on familiar attacks but suffer drastic recall drops on novel attacks. Unsupervised LOF attains moderate overall accuracy and high recall on unknown threats at the cost of elevated false alarms, while boundary-based OCSVM balances precision and recall best, demonstrating robust detection across both scenarios. These findings offer practical guidance for selecting IDS models in dynamic network environments.
Authors: Atharva Mehta, Shivam Chauhan, Monojit Choudhury
Abstract: Fine-tuning large-scale music generation models, such as MusicGen and Mustango, is a computationally expensive process, often requiring updates to billions of parameters and, therefore, significant hardware resources. Parameter-Efficient Fine-Tuning (PEFT) techniques, particularly adapter-based methods, have emerged as a promising alternative, enabling adaptation with minimal trainable parameters while preserving model performance. However, the design choices for adapters, including their architecture, placement, and size, are numerous, and it is unclear which of these combinations would produce optimal adapters and why, for a given case of low-resource music genre. In this paper, we attempt to answer this question by studying various adapter configurations for two AI music models, MusicGen and Mustango, on two genres: Hindustani Classical and Turkish Makam music. Our findings reveal distinct trade-offs: convolution-based adapters excel in capturing fine-grained local musical details such as ornamentations and short melodic phrases, while transformer-based adapters better preserve long-range dependencies crucial for structured improvisation. Additionally, we analyze computational resource requirements across different adapter scales, demonstrating how mid-sized adapters (40M parameters) achieve an optimal balance between expressivity and quality. Furthermore, we find that Mustango, a diffusion-based model, generates more diverse outputs with better adherence to the description in the input prompt while lacking in providing stability in notes, rhythm alignment, and aesthetics. Also, it is computationally intensive and requires significantly more time to train. In contrast, autoregressive models like MusicGen offer faster training and are more efficient, and can produce better quality output in comparison, but have slightly higher redundancy in their generations.
Authors: Sanghoon Yu, Min-hwan Oh
Abstract: We study the linear bandit problem under limited adaptivity, known as the batched linear bandit. While existing approaches can achieve near-optimal regret in theory, they are often computationally prohibitive or underperform in practice. We propose BLAE, a novel batched algorithm that integrates arm elimination with regularized G-optimal design, achieving the minimax optimal regret (up to logarithmic factors in $T$) in both large-$K$ and small-$K$ regimes for the first time, while using only $O(\log\log T)$ batches. Our analysis introduces new techniques for batch-wise optimal design and refined concentration bounds. Crucially, BLAE demonstrates low computational overhead and strong empirical performance, outperforming state-of-the-art methods in extensive numerical evaluations. Thus, BLAE is the first algorithm to combine provable minimax-optimality in all regimes and practical superiority in batched linear bandits.
Authors: Kun He
Abstract: The matrix scaling problem, particularly the Sinkhorn-Knopp algorithm, has been studied for over 60 years. In practice, the algorithm often yields high-quality approximations within just a few iterations. Theoretically, however, the best-known upper bound places it in the class of pseudopolynomial-time approximation algorithms. Meanwhile, the lower-bound landscape remains largely unexplored. Two fundamental questions persist: what accounts for the algorithm's strong empirical performance, and can a tight bound on its iteration count be established? For an $n\times n$ matrix, its normalized version is obtained by dividing each entry by its largest entry. We say that a normalized matrix has a density $\gamma$ if there exists a constant $\rho > 0$ such that one row or column has exactly $\lceil \gamma n \rceil$ entries with values at least $\rho$, and every other row and column has at least $\lceil \gamma n \rceil$ such entries. For the upper bound, we show that the Sinkhorn-Knopp algorithm produces a nearly doubly stochastic matrix in $O(\log n - \log \varepsilon)$ iterations and $\widetilde{O}(n^2)$ time for all nonnegative square matrices whose normalized version has a density $\gamma > 1/2$. Such matrices cover both the algorithm's principal practical inputs and its typical theoretical regime, and the $\widetilde{O}(n^2)$ runtime is optimal. For the lower bound, we establish a tight bound of $\widetilde{\Omega}\left(n^{1/2}/\varepsilon\right)$ iterations for positive matrices under the $\ell_2$-norm error measure. Moreover, for every $\gamma < 1/2$, there exists a matrix with density $\gamma$ for which the algorithm requires $\Omega\left(n^{1/2}/\varepsilon\right)$ iterations. In summary, our results reveal a sharp phase transition in the Sinkhorn-Knopp algorithm at the density threshold $\gamma = 1/2$.
Authors: Tao Tang, Chengxu Yang
Abstract: Pansharpening refers to the process of integrating a high resolution panchromatic (PAN) image with a lower resolution multispectral (MS) image to generate a fused product, which is pivotal in remote sensing. Despite the effectiveness of CNNs in addressing this challenge, they are inherently constrained by the uniform application of convolutional kernels across all spatial positions, overlooking local content variations. To overcome this issue, we introduce RAPNet, a new architecture that leverages content-adaptive convolution. At its core, RAPNet employs the Receptive-field Adaptive Pansharpening Convolution (RAPConv), designed to produce spatially adaptive kernels responsive to local feature context, thereby enhancing the precision of spatial detail extraction. Additionally, the network integrates the Pansharpening Dynamic Feature Fusion (PAN-DFF) module, which incorporates an attention mechanism to achieve an optimal balance between spatial detail enhancement and spectral fidelity. Comprehensive evaluations on publicly available datasets confirm that RAPNet delivers superior performance compared to existing approaches, as demonstrated by both quantitative metrics and qualitative assessments. Ablation analyses further substantiate the effectiveness of the proposed adaptive components.
Authors: Judy Long, Tao Liu, Sean Alexander Woznicki, Miljana Markovi\'c, Oskar Marko, Molly Sears
Abstract: Crop mapping involves identifying and classifying crop types using spatial data, primarily derived from remote sensing imagery. This study presents the first comprehensive review of large-scale, pixel-wise crop mapping workflows, encompassing both conventional supervised methods and emerging transfer learning approaches. To identify the optimal time-series generation approaches and supervised crop mapping models, we conducted systematic experiments, comparing six widely adopted satellite image-based preprocessing methods, alongside eleven supervised pixel-wise classification models. Additionally, we assessed the synergistic impact of varied training sample sizes and variable combinations. Moreover, we identified optimal transfer learning techniques for different magnitudes of domain shift. The evaluation of optimal methods was conducted across five diverse agricultural sites. Landsat 8 served as the primary satellite data source. Labels come from CDL trusted pixels and field surveys. Our findings reveal three key insights. First, fine-scale interval preprocessing paired with Transformer models consistently delivered optimal performance for both supervised and transferable workflows. RF offered rapid training and competitive performance in conventional supervised learning and direct transfer to similar domains. Second, transfer learning techniques enhanced workflow adaptability, with UDA being effective for homogeneous crop classes while fine-tuning remains robust across diverse scenarios. Finally, workflow choice depends heavily on the availability of labeled samples. With a sufficient sample size, supervised training typically delivers more accurate and generalizable results. Below a certain threshold, transfer learning that matches the level of domain shift is a viable alternative to achieve crop mapping. All code is publicly available to encourage reproducibility practice.
Authors: Xiaoran Xu, In-Ho Ra, Ravi Sankar
Abstract: Lung diseases, including lung cancer and COPD, are significant health concerns globally. Traditional diagnostic methods can be costly, time-consuming, and invasive. This study investigates the use of semi supervised learning methods for lung sound signal detection using a model combination of MFCC+CNN. By introducing semi supervised learning modules such as Mix Match, Co-Refinement, and Co Refurbishing, we aim to enhance the detection performance while reducing dependence on manual annotations. With the add-on semi-supervised modules, the accuracy rate of the MFCC+CNN model is 92.9%, an increase of 3.8% to the baseline model. The research contributes to the field of lung disease sound detection by addressing challenges such as individual differences, feature insufficient labeled data.
Authors: Bonwoo Lee, Cheolwoo Park, Jeongyoun Ahn
Abstract: Differential privacy has become a cornerstone in the development of privacy-preserving learning algorithms. This work addresses optimizing differentially private kernel learning within the empirical risk minimization (ERM) framework. We propose a novel differentially private kernel ERM algorithm based on random projection in the reproducing kernel Hilbert space using Gaussian processes. Our method achieves minimax-optimal excess risk for both the squared loss and Lipschitz-smooth convex loss functions under a local strong convexity condition. We further show that existing approaches based on alternative dimension reduction techniques, such as random Fourier feature mappings or $\ell_2$ regularization, yield suboptimal generalization performance. Our key theoretical contribution also includes the derivation of dimension-free generalization bounds for objective perturbation-based private linear ERM -- marking the first such result that does not rely on noisy gradient-based mechanisms. Additionally, we obtain sharper generalization bounds for existing differentially private kernel ERM algorithms. Empirical evaluations support our theoretical claims, demonstrating that random projection enables statistically efficient and optimally private kernel learning. These findings provide new insights into the design of differentially private algorithms and highlight the central role of dimension reduction in balancing privacy and utility.
Authors: Changxin Tian, Jiapeng Wang, Qian Zhao, Kunlong Chen, Jia Liu, Ziqi Liu, Jiaxin Mao, Wayne Xin Zhao, Zhiqiang Zhang, Jun Zhou
Abstract: Recent advances in learning rate (LR) scheduling have demonstrated the effectiveness of decay-free approaches that eliminate the traditional decay phase while maintaining competitive performance. Model merging techniques have emerged as particularly promising solutions in this domain. We present Warmup-Stable and Merge (WSM), a general framework that establishes a formal connection between learning rate decay and model merging. WSM provides a unified theoretical foundation for emulating various decay strategies-including cosine decay, linear decay and inverse square root decay-as principled model averaging schemes, while remaining fully compatible with diverse optimization methods. Through extensive experiments, we identify merge duration-the training window for checkpoint aggregation-as the most critical factor influencing model performance, surpassing the importance of both checkpoint interval and merge quantity. Our framework consistently outperforms the widely-adopted Warmup-Stable-Decay (WSD) approach across multiple benchmarks, achieving significant improvements of +3.5% on MATH, +2.9% on HumanEval, and +5.5% on MMLU-Pro. The performance advantages extend to supervised fine-tuning scenarios, highlighting WSM's potential for long-term model refinement.
Authors: Minghao Guo, Xi Zhu, Jingyuan Huang, Kai Mei, Yongfeng Zhang
Abstract: Graph Neural Networks (GNNs) have achieved remarkable success in graph-based learning by propagating information among neighbor nodes via predefined aggregation mechanisms. However, such fixed schemes often suffer from two key limitations. First, they cannot handle the imbalance in node informativeness -- some nodes are rich in information, while others remain sparse. Second, predefined message passing primarily leverages local structural similarity while ignoring global semantic relationships across the graph, limiting the model's ability to capture distant but relevant information. We propose Retrieval-augmented Graph Agentic Network (ReaGAN), an agent-based framework that empowers each node with autonomous, node-level decision-making. Each node acts as an agent that independently plans its next action based on its internal memory, enabling node-level planning and adaptive message propagation. Additionally, retrieval-augmented generation (RAG) allows nodes to access semantically relevant content and build global relationships in the graph. ReaGAN achieves competitive performance under few-shot in-context settings using a frozen LLM backbone without fine-tuning, showcasing the potential of agentic planning and local-global retrieval in graph learning.
Authors: Florian Felten, Gabriel Apaza, Gerhard Br\"aunlich, Cashen Diniz, Xuliang Dong, Arthur Drake, Milad Habibi, Nathaniel J. Hoffman, Matthew Keeler, Soheyl Massoudi, Francis G. VanGessel, Mark Fuge
Abstract: Engineering design optimization seeks to automatically determine the shapes, topologies, or parameters of components that maximize performance under given conditions. This process often depends on physics-based simulations, which are difficult to install, computationally expensive, and require domain-specific expertise. To mitigate these challenges, we introduce EngiBench, the first open-source library and datasets spanning diverse domains for data-driven engineering design. EngiBench provides a unified API and a curated set of benchmarks -- covering aeronautics, heat conduction, photonics, and more -- that enable fair, reproducible comparisons of optimization and machine learning algorithms, such as generative or surrogate models. We also release EngiOpt, a companion library offering a collection of such algorithms compatible with the EngiBench interface. Both libraries are modular, letting users plug in novel algorithms or problems, automate end-to-end experiment workflows, and leverage built-in utilities for visualization, dataset generation, feasibility checks, and performance analysis. We demonstrate their versatility through experiments comparing state-of-the-art techniques across multiple engineering design problems, an undertaking that was previously prohibitively time-consuming to perform. Finally, we show that these problems pose significant challenges for standard machine learning methods due to highly sensitive and constrained design manifolds.
Authors: Chunyu Liu, Hao Zhang, Wei Wu, Fuhui Zhou, Qihui Wu, Derrick Wing Kwan Ng, Chan-Byoung Chae
Abstract: The enhancement of spectrum efficiency and the realization of secure spectrum utilization are critically dependent on spectrum cognition. However, existing spectrum cognition methods often exhibit limited generalization and suboptimal accuracy when deployed across diverse spectrum environments and tasks. To overcome these challenges, we propose a spectrum foundation model, termed SpectrumFM, which provides a new paradigm for spectrum cognition. An innovative spectrum encoder that exploits the convolutional neural networks and the multi-head self attention mechanisms is proposed to effectively capture both fine-grained local signal structures and high-level global dependencies in the spectrum data. To enhance its adaptability, two novel self-supervised learning tasks, namely masked reconstruction and next-slot signal prediction, are developed for pre-training SpectrumFM, enabling the model to learn rich and transferable representations. Furthermore, low-rank adaptation (LoRA) parameter-efficient fine-tuning is exploited to enable SpectrumFM to seamlessly adapt to various downstream spectrum cognition tasks, including spectrum sensing (SS), anomaly detection (AD), and wireless technology classification (WTC). Extensive experiments demonstrate the superiority of SpectrumFM over state-of-the-art methods. Specifically, it improves detection probability in the SS task by 30% at -4 dB signal-to-noise ratio (SNR), boosts the area under the curve (AUC) in the AD task by over 10%, and enhances WTC accuracy by 9.6%.
Authors: Kirill Sokolov, Alexander Korotin
Abstract: We consider the discrete-time Schr\"odinger bridge problem on a finite state space. Although it has been known that the Iterative Markovian Fitting (IMF) algorithm converges in Kullback-Leibler divergence to the ground truth solution, the speed of that convergence remained unquantified. In this work, we establish for the first time that IMF exhibits exponential convergence with an explicit contraction factor.
Authors: Soumik Dey, Benjamin Braun, Naveen Ravipati, Hansi Wu, Binbin Li
Abstract: Sellers at eBay are recommended keyphrases to bid on to enhance the performance of their advertising campaigns. The relevance of these keyphrases is crucial in avoiding the overcrowding of search systems with irrelevant items and maintaining a positive seller perception. It is essential that keyphrase recommendations align with both seller and Search judgments regarding auctions. Due to the difficulty in procuring negative human judgment at scale, employing LLM-as-a-judge to mimic seller judgment has been established as the norm in several studies. This study introduces a novel two-step LLM distillation process from a LLM-judge used to debias our Embedding Based Retrieval (EBR) model from the various biases that exist in click-data. We distill from an LLM teacher via a cross-encoder assistant into a bi-encoder student using a multi-task training approach, ultimately employing the student bi-encoder to retrieve relevant advertiser keyphrases. We show that integrating a knowledge distillation process from LLMs in a multi-task training setup enhances bi-encoder performance in retrieving relevant advertiser keyphrases at eBay.
Authors: Lyria Team, Antoine Caillon, Brian McWilliams, Cassie Tarakajian, Ian Simon, Ilaria Manco, Jesse Engel, Noah Constant, Yunpeng Li, Timo I. Denk, Alberto Lalama, Andrea Agostinelli, Cheng-Zhi Anna Huang, Ethan Manilow, George Brower, Hakan Erdogan, Heidi Lei, Itai Rolnick, Ivan Grishchenko, Manu Orsini, Matej Kastelic, Mauricio Zuluaga, Mauro Verzetti, Michael Dooley, Ondrej Skopek, Rafael Ferrer, Zal\'an Borsos, \"Aaron van den Oord, Douglas Eck, Eli Collins, Jason Baldridge, Tom Hume, Chris Donahue, Kehang Han, Adam Roberts
Abstract: We introduce a new class of generative models for music called live music models that produce a continuous stream of music in real-time with synchronized user control. We release Magenta RealTime, an open-weights live music model that can be steered using text or audio prompts to control acoustic style. On automatic metrics of music quality, Magenta RealTime outperforms other open-weights music generation models, despite using fewer parameters and offering first-of-its-kind live generation capabilities. We also release Lyria RealTime, an API-based model with extended controls, offering access to our most powerful model with wide prompt coverage. These models demonstrate a new paradigm for AI-assisted music creation that emphasizes human-in-the-loop interaction for live music performance.
Authors: Bo Deng, Levi Heath
Abstract: In this paper, we describe a feedforward artificial neural network trained on the ImageNet 2012 contest dataset [7] with the new method of [5] to an accuracy rate of 98.3% with a 99.69 Top-1 rate, and an average of 285.9 labels that are perfectly classified over the 10 batch partitions of the dataset. The best performing model uses 322,430,160 parameters, with 4 decimal places precision. We conjecture that the reason our model does not achieve a 100% accuracy rate is due to a double-labeling problem, by which there are duplicate images in the dataset with different labels.
Authors: Bruno L. Pereira, Alan Said, Rodrygo L. T. Santos
Abstract: Offline evaluation plays a central role in benchmarking recommender systems when online testing is impractical or risky. However, it is susceptible to two key sources of bias: exposure bias, where users only interact with items they are shown, and sampling bias, introduced when evaluation is performed on a subset of logged items rather than the full catalog. While prior work has proposed methods to mitigate sampling bias, these are typically assessed on fixed logged datasets rather than for their ability to support reliable model comparisons under varying exposure conditions or relative to true user preferences. In this paper, we investigate how different combinations of logging and sampling choices affect the reliability of offline evaluation. Using a fully observed dataset as ground truth, we systematically simulate diverse exposure biases and assess the reliability of common sampling strategies along four dimensions: sampling resolution (recommender model separability), fidelity (agreement with full evaluation), robustness (stability under exposure bias), and predictive power (alignment with ground truth). Our findings highlight when and how sampling distorts evaluation outcomes and offer practical guidance for selecting strategies that yield faithful and robust offline comparisons.
Authors: Andrew Cropper, David M. Cerna, Matti J\"arvisalo
Abstract: The goal of inductive logic programming is to search for a hypothesis that generalises training data and background knowledge. The challenge is searching vast hypothesis spaces, which is exacerbated because many logically equivalent hypotheses exist. To address this challenge, we introduce a method to break symmetries in the hypothesis space. We implement our idea in answer set programming. Our experiments on multiple domains, including visual reasoning and game playing, show that our approach can reduce solving times from over an hour to just 17 seconds.