Authors: Antonio-Gabriel Chac\'on Menke (Shibaura Institute of Technology, Kempten University of Applied Sciences), Phan Xuan Tan (Shibaura Institute of Technology)
Abstract: Recent incidents highlight safety risks in Large Language Models (LLMs), motivating research into alignment methods like Constitutional AI (CAI). This paper explores CAI's self-critique mechanism on small, uncensored 7-9B parameter models: DeepSeek-R1, Gemma-2, Llama 3.1, and Qwen2.5. Using HarmBench, we demonstrate that while all models showed capacity for harm reduction through self-critique, effectiveness varied significantly, with DeepSeek-R1's explicit reasoning process yielding superior results. These findings suggest that CAI-inspired prompting strategies can enhance safety in resource-constrained models, though success depends on the model's capacity for harm detection.
Authors: Andrew Kiruluta, Andreas Lemos
Abstract: In recent years, diffusion based methods have emerged as a powerful paradigm for generative modeling. Although discrete diffusion for natural language processing has been explored to a lesser extent, it shows promise for tasks requiring iterative denoising of token based data. In standard approaches to text generation, transformers dominate, but their reliance on self attention often incurs high computational costs. This paper introduces a fully diffusion driven discrete text generation model built without any transformer or large convolution modules. Instead, the model integrates structured state space dynamics in the time domain with a novel Complex Fourier Multi Layer Perceptron module that operates in the frequency domain. The forward noising process randomly samples the vocabulary to replace tokens with a controlled probability, while the learned reverse model systematically reverts corrupted sequences toward their original states. By composing local state space updates with global Fourier based mixing, the approach effectively captures both short and long range dependencies.
Authors: Subed Lamichhane, Haotian Lu, Sheldon X. -D. Tan
Abstract: In contrast to the assumptions of most existing Electromigration (EM) analysis tools, the evolution of EM-induced stress is inherently non-deterministic, influenced by factors such as input current fluctuations and manufacturing non-idealities. Traditional approaches for estimating stress variations typically involve computationally expensive and inefficient Monte Carlo simulations with industrial solvers, which quantify variations using mean and variance metrics. In this work, we introduce a novel machine learning-based framework, termed BPINNEM- Post, for efficient stochastic analysis of EM-induced postvoiding aging processes. This new approach integrates closedform analytical solutions with a Bayesian Physics-Informed Neural Network (BPINN) framework to accelerate the analysis for the first time. The closed-form solutions enforce physical laws at the individual wire segment level, while the BPINN ensures that physics constraints at inter-segment junctions are satisfied and stochastic behaviors are accurately modeled. By reducing the number of variables in the loss functions through the use of analytical solutions, our method significantly improves training efficiency without accuracy loss and naturally incorporates variational effects. Additionally, the analytical solutions effectively address the challenge of incorporating initial stress distributions in interconnect structures during post-void stress calculations. Numerical results demonstrate that BPINN-EM-Post achieves over 240x speedup compared to Monte Carlo simulations using the FEM-based COMSOL solver and more than 65x speedup compared to Monte Carlo simulations using the FDM-based EMSpice method.
Authors: Kangrui Du, Yuhang Wu, Shikuang Deng, Shi Gu
Abstract: Spiking Neural Networks (SNNs), models inspired by neural mechanisms in the brain, allow for energy-efficient implementation on neuromorphic hardware. However, SNNs trained with current direct training approaches are constrained to a specific time step. This "temporal inflexibility" 1) hinders SNNs' deployment on time-step-free fully event-driven chips and 2) prevents energy-performance balance based on dynamic inference time steps. In this study, we first explore the feasibility of training SNNs that generalize across different time steps. We then introduce Mixed Time-step Training (MTT), a novel method that improves the temporal flexibility of SNNs, making SNNs adaptive to diverse temporal structures. During each iteration of MTT, random time steps are assigned to different SNN stages, with spikes transmitted between stages via communication modules. After training, the weights are deployed and evaluated on both time-stepped and fully event-driven platforms. Experimental results show that models trained by MTT gain remarkable temporal flexibility, friendliness for both event-driven and clock-driven deployment (nearly lossless on N-MNIST and 10.1% higher than standard methods on CIFAR10-DVS), enhanced network generalization, and near SOTA performance. To the best of our knowledge, this is the first work to report the results of large-scale SNN deployment on fully event-driven scenarios.
Authors: Oscar L. Cruz-Gonz\'alez, Val\'erie Deplano, Badih Ghattas
Abstract: Due to the limited accuracy of 4D Magnetic Resonance Imaging (MRI) in identifying hemodynamics in cardiovascular diseases, the challenges in obtaining patient-specific flow boundary conditions, and the computationally demanding and time-consuming nature of Computational Fluid Dynamics (CFD) simulations, it is crucial to explore new data assimilation algorithms that offer possible alternatives to these limitations. In the present work, we study Physics-Informed Neural Networks (PINNs), Deep Operator Networks (DeepONets), and their Physics-Informed extensions (PI-DeepONets) in predicting vascular flow simulations in the context of a 3D Abdominal Aortic Aneurysm (AAA) idealized model. PINN is a technique that combines deep neural networks with the fundamental principles of physics, incorporating the physics laws, which are given as partial differential equations, directly into loss functions used during the training process. On the other hand, DeepONet is designed to learn nonlinear operators from data and is particularly useful in studying parametric partial differential equations (PDEs), e.g., families of PDEs with different source terms, boundary conditions, or initial conditions. Here, we adapt the approaches to address the particular use case of AAA by integrating the 3D Navier-Stokes equations (NSE) as the physical laws governing fluid dynamics. In addition, we follow best practices to enhance the capabilities of the models by effectively capturing the underlying physics of the problem under study. The advantages and limitations of each approach are highlighted through a series of relevant application cases. We validate our results by comparing them with CFD simulations for benchmark datasets, demonstrating good agreements and emphasizing those cases where improvements in computational efficiency are observed.
Authors: Maisha Binte Rashid, Pablo Rivas
Abstract: In this paper, we aim to investigate the capabilities of multimodal machine learning models, particularly the OpenFlamingo model, in processing a large-scale dataset of consumer-to-consumer (C2C) online posts related to car parts. We have collected data from two platforms, OfferUp and Craigslist, resulting in a dataset of over 1.2 million posts with their corresponding images. The OpenFlamingo model was used to extract embeddings for the text and image of each post. We used $k$-means clustering on the joint embeddings to identify underlying patterns and commonalities among the posts. We have found that most clusters contain a pattern, but some clusters showed no internal patterns. The results provide insight into the fact that OpenFlamingo can be used for finding patterns in large datasets but needs some modification in the architecture according to the dataset.
Authors: Minheng Xiao, Zhenbang Jiao
Abstract: In many practical reinforcement learning scenarios, feedback is provided only at the end of a long horizon, leading to sparse and delayed rewards. Existing reward redistribution methods typically assume that per-step rewards are independent, thus overlooking interdependencies among state--action pairs. In this paper, we propose a \emph{Likelihood Reward Redistribution} (LRR) framework that addresses this issue by modeling each per-step reward with a parametric probability distribution whose parameters depend on the state--action pair. By maximizing the likelihood of the observed episodic return via a leave-one-out (LOO) strategy that leverages the entire trajectory, our framework inherently introduces an uncertainty regularization term into the surrogate objective. Moreover, we show that the conventional mean squared error (MSE) loss for reward redistribution emerges as a special case of our likelihood framework when the uncertainty is fixed under the Gaussian distribution. When integrated with an off-policy algorithm such as Soft Actor-Critic, LRR yields dense and informative reward signals, resulting in superior sample efficiency and policy performance on Box-2d and MuJoCo benchmarks.
Authors: Josef Koumar, Timotej Smole\v{n}, Kamil Je\v{r}\'abek, Tom\'a\v{s} \v{C}ejka
Abstract: Accurate network traffic forecasting is essential for Internet Service Providers (ISP) to optimize resources, enhance user experience, and mitigate anomalies. This study evaluates state-of-the-art deep learning models on CESNET-TimeSeries24, a recently published, comprehensive real-world network traffic dataset from the ISP network CESNET3 spanning multivariate time series over 40 weeks. Our findings highlight the balance between prediction accuracy and computational efficiency across different levels of network granularity. Additionally, this work establishes a reproducible methodology that facilitates direct comparison of existing approaches, explores their strengths and weaknesses, and provides a benchmark for future studies using this dataset.
Authors: Jungkyoo Shin, Bumsoo Kim, Eunwoo Kim
Abstract: Multi-modal understanding plays a crucial role in artificial intelligence by enabling models to jointly interpret inputs from different modalities. However, conventional approaches such as contrastive learning often struggle with modality discrepancies, leading to potential misalignments. In this paper, we propose a novel class anchor alignment approach that leverages class probability distributions for multi-modal representation learning. Our method, Class-anchor-ALigned generative Modeling (CALM), encodes class anchors as prompts to generate and align class probability distributions for each modality, enabling more effective alignment. Furthermore, we introduce a cross-modal probabilistic variational autoencoder to model uncertainty in the alignment, enhancing the ability to capture deeper relationships between modalities and data variations. Extensive experiments on four benchmark datasets demonstrate that our approach significantly outperforms state-of-the-art methods, especially in out-of-domain evaluations. This highlights its superior generalization capabilities in multi-modal representation learning.
Authors: Javier J. Poveda Rodrigo, Mohamed Amine Ahmdi, Alessio Burrello, Daniele Jahier Pagliari, Luca Benini
Abstract: The recent exponential growth of Large Language Models (LLMs) has relied on GPU-based systems. However, CPUs are emerging as a flexible and lower-cost alternative, especially when targeting inference and reasoning workloads. RISC-V is rapidly gaining traction in this area, given its open and vendor-neutral ISA. However, the RISC-V hardware for LLM workloads and the corresponding software ecosystem are not fully mature and streamlined, given the requirement of domain-specific tuning. This paper aims at filling this gap, focusing on optimizing LLM inference on the Sophon SG2042, the first commercially available many-core RISC-V CPU with vector processing capabilities. On two recent state-of-the-art LLMs optimized for reasoning, DeepSeek R1 Distill Llama 8B and DeepSeek R1 Distill QWEN 14B, we achieve 4.32/2.29 token/s for token generation and 6.54/3.68 token/s for prompt processing, with a speed up of up 2.9x/3.0x compared to our baseline.
Authors: Cyrus Malik, Josef Bajada, Joshua Ellul
Abstract: The evaluation of smart contract reputability is essential to foster trust in decentralized ecosystems. However, existing methods that rely solely on static code analysis or transactional data, offer limited insight into evolving trustworthiness. We propose a multimodal data fusion framework that integrates static code features with transactional data to enhance reputability prediction. Our framework initially focuses on static code analysis, utilizing GAN-augmented opcode embeddings to address class imbalance, achieving 97.67% accuracy and a recall of 0.942 in detecting illicit contracts, surpassing traditional oversampling methods. This forms the crux of a reputability-centric fusion strategy, where combining static and transactional data improves recall by 7.25% over single-source models, demonstrating robust performance across validation sets. By providing a holistic view of smart contract behaviour, our approach enhances the model's ability to assess reputability, identify fraudulent activities, and predict anomalous patterns. These capabilities contribute to more accurate reputability assessments, proactive risk mitigation, and enhanced blockchain security.
Authors: Lars Kr\"oger, Cristian Cioflan, Victor Kartsch, Luca Benini
Abstract: RISC-V-based architectures are paving the way for efficient On-Device Learning (ODL) in smart edge devices. When applied across multiple nodes, ODL enables the creation of intelligent sensor networks that preserve data privacy. However, developing ODL-capable, battery-operated embedded platforms presents significant challenges due to constrained computational resources and limited device lifetime, besides intrinsic learning issues such as catastrophic forgetting. We face these challenges by proposing a regularization-based On-Device Federated Continual Learning algorithm tailored for multiple nano-drones performing face recognition tasks. We demonstrate our approach on a RISC-V-based 10-core ultra-low-power SoC, optimizing the ODL computational requirements. We improve the classification accuracy by 24% over naive fine-tuning, requiring 178 ms per local epoch and 10.5 s per global epoch, demonstrating the effectiveness of the architecture for this task.
Authors: Zhuoshi Pan, Yu Li, Honglin Lin, Qizhi Pei, Zinan Tang, Wei Wu, Chenlin Ming, H. Vicky Zhao, Conghui He, Lijun Wu
Abstract: Large language models (LLMs) have demonstrated remarkable reasoning capability in solving mathematical problems. However, existing approaches primarily focus on improving the quality of correct training data, e.g., distilling high-quality correct solutions from advanced models, neglecting the value contained in error data, potentially hindering the model's reflective ability. Though some studies attempt to leverage error data, they often involve complex mechanisms, such as Monte Carlo Tree Search (MCTS) to explore error nodes. In this work, we propose to enhance LLMs' reasoning ability by Learning from Errors for Mathematical Advancement (LEMMA). LEMMA constructs data consisting of an incorrect solution with an erroneous step and a reflection connection to a correct solution for fine-tuning. Specifically, we systematically analyze the model-generated error types and introduce an error-type grounded mistake augmentation method to collect diverse and representative errors. Correct solutions are either from fixing the errors or generating a fresh start. Through a model-aware smooth reflection connection, the erroneous solution is transferred to the correct one. By fine-tuning on the constructed dataset, the model is able to self-correct errors autonomously within the generation process without relying on external critique models. Experimental results demonstrate that LEMMA achieves significant performance improvements over other strong baselines.
Authors: Gideon Stein, Maha Shadaydeh, Jan Blunk, Niklas Penzel, Joachim Denzler
Abstract: Causal discovery, or identifying causal relationships from observational data, is a notoriously challenging task, with numerous methods proposed to tackle it. Despite this, in-the-wild evaluation of these methods is still lacking, as works frequently rely on synthetic data evaluation and sparse real-world examples under critical theoretical assumptions. Real-world causal structures, however, are often complex, making it hard to decide on a proper causal discovery strategy. To bridge this gap, we introduce CausalRivers, the largest in-the-wild causal discovery benchmarking kit for time-series data to date. CausalRivers features an extensive dataset on river discharge that covers the eastern German territory (666 measurement stations) and the state of Bavaria (494 measurement stations). It spans the years 2019 to 2023 with a 15-minute temporal resolution. Further, we provide additional data from a flood around the Elbe River, as an event with a pronounced distributional shift. Leveraging multiple sources of information and time-series meta-data, we constructed two distinct causal ground truth graphs (Bavaria and eastern Germany). These graphs can be sampled to generate thousands of subgraphs to benchmark causal discovery across diverse and challenging settings. To demonstrate the utility of CausalRivers, we evaluate several causal discovery approaches through a set of experiments to identify areas for improvement. CausalRivers has the potential to facilitate robust evaluations and comparisons of causal discovery methods. Besides this primary purpose, we also expect that this dataset will be relevant for connected areas of research, such as time-series forecasting and anomaly detection. Based on this, we hope to push benchmark-driven method development that fosters advanced techniques for causal discovery, as is the case for many other areas of machine learning.
Authors: Ali Beikmohammadi, Sarit Khirirat, Peter Richt\'arik, Sindri Magn\'usson
Abstract: Federated reinforcement learning (FedRL) enables collaborative learning while preserving data privacy by preventing direct data exchange between agents. However, many existing FedRL algorithms assume that all agents operate in identical environments, which is often unrealistic. In real-world applications -- such as multi-robot teams, crowdsourced systems, and large-scale sensor networks -- each agent may experience slightly different transition dynamics, leading to inherent model mismatches. In this paper, we first establish linear convergence guarantees for single-agent temporal difference learning (TD(0)) in policy evaluation and demonstrate that under a perturbed environment, the agent suffers a systematic bias that prevents accurate estimation of the true value function. This result holds under both i.i.d. and Markovian sampling regimes. We then extend our analysis to the federated TD(0) (FedTD(0)) setting, where multiple agents -- each interacting with its own perturbed environment -- periodically share value estimates to collaboratively approximate the true value function of a common underlying model. Our theoretical results indicate the impact of model mismatch, network connectivity, and mixing behavior on the convergence of FedTD(0). Empirical experiments corroborate our theoretical gains, highlighting that even moderate levels of information sharing can significantly mitigate environment-specific errors.
Authors: Sahil Tyagi, Prateek Sharma
Abstract: Deep learning systems are optimized for clusters with homogeneous resources. However, heterogeneity is prevalent in computing infrastructure across edge, cloud and HPC. When training neural networks using stochastic gradient descent techniques on heterogeneous resources, performance degrades due to stragglers and stale updates. In this work, we develop an adaptive batch-scaling framework called OmniLearn to mitigate the effects of heterogeneity in distributed training. Our approach is inspired by proportional controllers to balance computation across heterogeneous servers, and works under varying resource availability. By dynamically adjusting worker mini-batches at runtime, OmniLearn reduces training time by 14-85%. We also investigate asynchronous training, where our techniques improve accuracy by up to 6.9%.
Authors: Miguel L\'opez-P\'erez, Marco Miani, Valery Naranjo, S{\o}ren Hauberg, Aasa Feragen
Abstract: Generative models are popular for medical imaging tasks such as anomaly detection, feature extraction, data visualization, or image generation. Since they are parameterized by deep learning models, they are often sensitive to distribution shifts and unreliable when applied to out-of-distribution data, creating a risk of, e.g. underrepresentation bias. This behavior can be flagged using uncertainty quantification methods for generative models, but their availability remains limited. We propose SLUG: A new UQ method for VAEs that combines recent advances in Laplace approximations with stochastic trace estimators to scale gracefully with image dimensionality. We show that our UQ score -- unlike the VAE's encoder variances -- correlates strongly with reconstruction error and racial underrepresentation bias for dermatological images. We also show how pixel-wise uncertainty can detect out-of-distribution image content such as ink, rulers, and patches, which is known to induce learning shortcuts in predictive models.
Authors: Keyon Vafa, Sarah Bentley, Jon Kleinberg, Sendhil Mullainathan
Abstract: How should we evaluate the quality of generative models? Many existing metrics focus on a model's producibility, i.e. the quality and breadth of outputs it can generate. However, the actual value from using a generative model stems not just from what it can produce but whether a user with a specific goal can produce an output that satisfies that goal. We refer to this property as steerability. In this paper, we first introduce a mathematical framework for evaluating steerability independently from producibility. Steerability is more challenging to evaluate than producibility because it requires knowing a user's goals. We address this issue by creating a benchmark task that relies on one key idea: sample an output from a generative model and ask users to reproduce it. We implement this benchmark in a large-scale user study of text-to-image models and large language models. Despite the ability of these models to produce high-quality outputs, they all perform poorly on steerabilty. This suggests that we need to focus on improving the steerability of generative models. We show such improvements are indeed possible: through reinforcement learning techniques, we create an alternative steering mechanism for image models that achieves more than 2x improvement on this benchmark.
Authors: Shivam Gupta, Sushrut Karmalkar
Abstract: Knowledge distillation is a technique used to train a small student network using the output generated by a large teacher network, and has many empirical advantages~\citep{Hinton2015DistillingTK}. While the standard one-shot approach to distillation only uses the output of the final teacher network, recent work~\citep{panigrahi2024progressive} has shown that using intermediate checkpoints from the teacher's training process as an implicit ``curriculum'' for progressive distillation can significantly speed up training. However, such schemes require storing these checkpoints, and often require careful selection of the intermediate checkpoints to train on, which can be impractical for large-scale training. In this paper, we show that a curriculum can be \emph{extracted} from just the fully trained teacher network, and that this extracted curriculum can give similar efficiency benefits to those of progressive distillation. Our extraction scheme is natural; we use a random projection of the hidden representations of the teacher network to progressively train the student network, before training using the output of the full network. We show that our scheme significantly outperforms one-shot distillation and achieves a performance similar to that of progressive distillation for learning sparse parities with two-layer networks, and provide theoretical guarantees for this setting. Additionally, we show that our method outperforms one-shot distillation even when using transformer-based architectures, both for sparse-parity learning, and language modeling tasks.
Authors: Louis Owen, Abhay Kumar, Nilabhra Roy Chowdhury, Fabian G\"ura
Abstract: The outcome of Large Language Model (LLM) pre-training strongly depends on weight initialization and variance control strategies. Although the importance of initial variance control has been well documented in neural networks in general, the literature on initialization and management of its growth during LLM pre-training, specifically, is somewhat sparse. In this paper, we introduce the Layer Index Rescaling (LIR) weight initialization scheme, and the Target Variance Rescaling (TVR) variance control strategy. Experiments on a 1B parameter LLaMA model demonstrate that better variance management using these techniques yields substantial improvements in downstream task performance (up to 4.6% on common pre-training benchmarks) and reduces extreme activation values, thus mitigating challenges associated with quantization and low-precision training. Our code is available at: https://github.com/bluorion-com/weight_rescaling.
Authors: Anran Xu, Lindsey J. Heagy
Abstract: In this work, we employ neural fields, which use neural networks to map a coordinate to the corresponding physical property value at that coordinate, in a test-time learning manner. For a test-time learning method, the weights are learned during the inversion, as compared to traditional approaches which require a network to be trained using a training data set. Results for synthetic examples in seismic tomography and direct current resistivity inversions are shown first. We then perform a singular value decomposition analysis on the Jacobian of the weights of the neural network (SVD analysis) for both cases to explore the effects of neural networks on the recovered model. The results show that the test-time learning approach can eliminate unwanted artifacts in the recovered subsurface physical property model caused by the sensitivity of the survey and physics. Therefore, NFs-Inv improves the inversion results compared to the conventional inversion in some cases such as the recovery of the dip angle or the prediction of the boundaries of the main target. In the SVD analysis, we observe similar patterns in the left-singular vectors as were observed in some diffusion models, trained in a supervised manner, for generative tasks in computer vision. This observation provides evidence that there is an implicit bias, which is inherent in neural network structures, that is useful in supervised learning and test-time learning models. This implicit bias has the potential to be useful for recovering models in geophysical inversions.
Authors: Navaneeth N, Souvik Chakraborty
Abstract: Modeling cardiovascular anatomies poses a significant challenge due to their complex, irregular structures and inherent pathological conditions. Numerical simulations, while accurate, are often computationally expensive, limiting their practicality in clinical settings. Traditional machine learning methods, on the other hand, often struggle with some major hurdles, including high dimensionality of the inputs, inability to effectively work with irregular grids, and preserving the time dependencies of responses in dynamic problems. In response to these challenges, we propose a geometry adaptive waveformer model to predict blood flow dynamics in the cardiovascular system. The framework is primarily composed of three components: a geometry encoder, a geometry decoder, and a waveformer. The encoder transforms input defined on the irregular domain to a regular domain using a graph operator-based network and signed distance functions. The waveformer operates on the transformed field on the irregular grid. Finally, the decoder reverses this process, transforming the output from the regular grid back to the physical space. We evaluate the efficacy of the approach on different sets of cardiovascular data.
Authors: Giuseppe Franco, Pablo Monteagudo-Lago, Ian Colbert, Nicholas Fraser, Michaela Blott
Abstract: The size of a model has been a strong predictor of its quality, as well as its cost. As such, the trade-off between model cost and quality has been well-studied. Post-training optimizations like quantization and pruning have typically focused on reducing the overall volume of pre-trained models to reduce inference costs while maintaining model quality. However, recent advancements have introduced optimization techniques that, interestingly, expand models post-training, increasing model size to improve quality when reducing volume. For instance, to enable 4-bit weight and activation quantization, incoherence processing often necessitates inserting online Hadamard rotations in the compute graph, and preserving highly sensitive weights often calls for additional higher precision computations. However, if application requirements cannot be met, the prevailing solution is to relax quantization constraints. In contrast, we demonstrate post-training model expansion is a viable strategy to improve model quality within a quantization co-design space, and provide theoretical justification. We show it is possible to progressively and selectively expand the size of a pre-trained large language model (LLM) to improve model quality without end-to-end retraining. In particular, when quantizing the weights and activations to 4 bits for Llama3 1B, we reduce the zero-shot accuracy gap to full precision by an average of 3% relative to both QuaRot and SpinQuant with only 5% more parameters, which is still a 3.8% reduction in volume relative to a BF16 reference model.
Authors: \'Italo Romani de Oliveira, Samet Ayhan, Glaucia Balvedi, Michael Biglin, Pablo Costas, Euclides C. Pinto Neto, Alexandre Leite, Felipe C. F. de Azevedo
Abstract: Predicting air traffic congestion and flow management is essential for airlines and Air Navigation Service Providers (ANSP) to enhance operational efficiency. Accurate estimates of future airport capacity and airspace density are vital for better airspace management, reducing air traffic controller workload and fuel consumption, ultimately promoting sustainable aviation. While existing literature has addressed these challenges, data management and query processing remain complex due to the vast volume of high-rate air traffic data. Many analytics use cases require a common pre-processing infrastructure, as ad-hoc approaches are insufficient. Additionally, linear prediction models often fall short, necessitating more advanced techniques. This paper presents a data processing and predictive services architecture that ingests large, uncorrelated, and noisy streaming data to forecast future airspace system states. The system continuously collects raw data, periodically compresses it, and stores it in NoSQL databases for efficient query processing. For prediction, the system learns from historical traffic by extracting key features such as airport arrival and departure events, sector boundary crossings, weather parameters, and other air traffic data. These features are input into various regression models, including linear, non-linear, and ensemble models, with the best-performing model selected for predictions. We evaluate this infrastructure across three prediction use cases in the US National Airspace System (NAS) and a segment of European airspace, using extensive real operations data, confirming that our system can predict future system states efficiently and accurately.
Authors: Amin Abbasishahkoo, Mahboubeh Dadkhah, Lionel Briand, Dayi Lin
Abstract: Deep Neural Networks (DNNs) face challenges during deployment due to data distribution shifts. Fine-tuning adapts pre-trained models to new contexts requiring smaller labeled sets. However, testing fine-tuned models under constrained labeling budgets remains a critical challenge. This paper introduces MetaSel, a new approach, tailored for fine-tuned DNN models, to select tests from unlabeled inputs. MetaSel assumes that fine-tuned and pre-trained models share related data distributions and exhibit similar behaviors for many inputs. However, their behaviors diverge within the input subspace where fine-tuning alters decision boundaries, making those inputs more prone to misclassification. Unlike general approaches that rely solely on the DNN model and its input set, MetaSel leverages information from both the fine-tuned and pre-trained models and their behavioral differences to estimate misclassification probability for unlabeled test inputs, enabling more effective test selection. Our extensive empirical evaluation, comparing MetaSel against 10 state-of-the-art approaches and involving 68 fine-tuned models across weak, medium, and strong distribution shifts, demonstrates that MetaSel consistently delivers significant improvements in Test Relative Coverage (TRC) over existing baselines, particularly under highly constrained labeling budgets. MetaSel shows average TRC improvements of 28.46% to 56.18% over the most frequent second-best baselines while maintaining a high TRC median and low variability. Our results confirm MetaSel's practicality, robustness, and cost-effectiveness for test selection in the context of fine-tuned models.
Authors: Bart Bussmann, Noa Nabeshima, Adam Karvonen, Neel Nanda
Abstract: Sparse autoencoders (SAEs) have emerged as a powerful tool for interpreting neural networks by extracting the concepts represented in their activations. However, choosing the size of the SAE dictionary (i.e. number of learned concepts) creates a tension: as dictionary size increases to capture more relevant concepts, sparsity incentivizes features to be split or absorbed into more specific features, leaving high-level features missing or warped. We introduce Matryoshka SAEs, a novel variant that addresses these issues by simultaneously training multiple nested dictionaries of increasing size, forcing the smaller dictionaries to independently reconstruct the inputs without using the larger dictionaries. This organizes features hierarchically - the smaller dictionaries learn general concepts, while the larger dictionaries learn more specific concepts, without incentive to absorb the high-level features. We train Matryoshka SAEs on Gemma-2-2B and TinyStories and find superior performance on sparse probing and targeted concept erasure tasks, more disentangled concept representations, and reduced feature absorption. While there is a minor tradeoff with reconstruction performance, we believe Matryoshka SAEs are a superior alternative for practical tasks, as they enable training arbitrarily large SAEs while retaining interpretable features at different levels of abstraction.
Authors: Khadija Zanna, Akane Sano
Abstract: Causal discovery (CD) plays a pivotal role in numerous scientific fields by clarifying the causal relationships that underlie phenomena observed in diverse disciplines. Despite significant advancements in CD algorithms that enhance bias and fairness analyses in machine learning, their application faces challenges due to the high computational demands and complexities of large-scale data. This paper introduces a framework that leverages Large Language Models (LLMs) for CD, utilizing a metadata-based approach akin to the reasoning processes of human experts. By shifting from pairwise queries to a more scalable breadth-first search (BFS) strategy, the number of required queries is reduced from quadratic to linear in terms of variable count, thereby addressing scalability concerns inherent in previous approaches. This method utilizes an Active Learning (AL) and a Dynamic Scoring Mechanism that prioritizes queries based on their potential information gain, combining mutual information, partial correlation, and LLM confidence scores to refine the causal graph more efficiently and accurately. This BFS query strategy reduces the required number of queries significantly, thereby addressing scalability concerns inherent in previous approaches. This study provides a more scalable and efficient solution for leveraging LLMs in fairness-driven CD, highlighting the effects of the different parameters on performance. We perform fairness analyses on the inferred causal graphs, identifying direct and indirect effects of sensitive attributes on outcomes. A comparison of these analyses against those from graphs produced by baseline methods highlights the importance of accurate causal graph construction in understanding bias and ensuring fairness in machine learning systems.
Authors: Victor Ulisses Pugliese, Os\'eias F. de A. Ferreira, Fabio A. Faria
Abstract: This paper proposes a novel approach based on deep reinforcement learning (DRL) for the 2D+1 packing problem with spatial constraints. This problem is an extension of the traditional 2D packing problem, incorporating an additional constraint on the height dimension. Therefore, a simulator using the OpenAI Gym framework has been developed to efficiently simulate the packing of rectangular pieces onto two boards with height constraints. Furthermore, the simulator supports multidiscrete actions, enabling the selection of a position on either board and the type of piece to place. Finally, two DRL-based methods (Proximal Policy Optimization -- PPO and the Advantage Actor-Critic -- A2C) have been employed to learn a packing strategy and demonstrate its performance compared to a well-known heuristic baseline (MaxRect-BL). In the experiments carried out, the PPO-based approach proved to be a good solution for solving complex packaging problems and highlighted its potential to optimize resource utilization in various industrial applications, such as the manufacturing of aerospace composites.
Authors: Sharon Lin (Dj), Krishnamurthy (Dj), Dvijotham, Jamie Hayes, Chongyang Shi, Ilia Shumailov, Shuang Song
Abstract: Backdoor attacks on machine learning models have been extensively studied, primarily within the computer vision domain. Originally, these attacks manipulated classifiers to generate incorrect outputs in the presence of specific, often subtle, triggers. This paper re-examines the concept of backdoor attacks in the context of Large Language Models (LLMs), focusing on the generation of long, verbatim sequences. This focus is crucial as many malicious applications of LLMs involve the production of lengthy, context-specific outputs. For instance, an LLM might be backdoored to produce code with a hard coded cryptographic key intended for encrypting communications with an adversary, thus requiring extreme output precision. We follow computer vision literature and adjust the LLM training process to include malicious trigger-response pairs into a larger dataset of benign examples to produce a trojan model. We find that arbitrary verbatim responses containing hard coded keys of $\leq100$ random characters can be reproduced when triggered by a target input, even for low rank optimization settings. Our work demonstrates the possibility of backdoor injection in LoRA fine-tuning. Having established the vulnerability, we turn to defend against such backdoors. We perform experiments on Gemini Nano 1.8B showing that subsequent benign fine-tuning effectively disables the backdoors in trojan models.
Authors: Jaeyeon Lee, Guantong Qi, Matthew Brady Neeley, Zhandong Liu, Hyun-Hwan Jeong
Abstract: Recent advancements in large language models (LLMs) integrating explicit reasoning, such as OpenAI's o3-mini, DeepSeek-R1, and QWQ-32B, enable smaller models to solve complex tasks by generating intermediate reasoning steps prior to providing answers. However, this approach significantly increases computational costs, both monetarily and environmentally. The widely-used self-consistency method further exacerbates these costs by aggregating multiple reasoning paths to improve accuracy, often requiring between 40 to 64 samples per task. Although aggregation effectively reduces variance and bias, additional sampling can lead to diminishing returns when early samples yield consistent results. To address inefficiencies, we propose leveraging Sequential Probability Ratio Testing (SPRT) to dynamically terminate sampling once sufficient consistency is achieved. We calibrate SPRT parameters specifically for LLM applications, accounting for sensitivity to detect the mode of the distribution. Our experiments demonstrate that incorporating SPRT significantly enhances token efficiency, achieving comparable accuracy to self-consistency methods but at a substantially reduced computational cost. To promote transparency and facilitate reproducibility, we have made the source code and datasets used in our experiments publicly available at our GitHub repository: https://github.com/LiuzLab/consol, or available as a PyPI package: pip install consol. We hope that this resource will support further research and encourage the development of new methods building upon our work.
Authors: Methusela Sulle, Judith Mwakalonge, Gurcan Comert, Saidi Siuhi, Nana Kankam Gyimah
Abstract: Road fatalities pose significant public safety and health challenges worldwide, with pedestrians being particularly vulnerable in vehicle-pedestrian crashes due to disparities in physical and performance characteristics. This study employs explainable artificial intelligence (XAI) to identify key factors contributing to pedestrian fatalities across the five U.S. states with the highest crash rates (2018-2022). It compares them to the five states with the lowest fatality rates. Using data from the Fatality Analysis Reporting System (FARS), the study applies machine learning techniques-including Decision Trees, Gradient Boosting Trees, Random Forests, and XGBoost-to predict contributing factors to pedestrian fatalities. To address data imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) is utilized, while SHapley Additive Explanations (SHAP) values enhance model interpretability. The results indicate that age, alcohol and drug use, location, and environmental conditions are significant predictors of pedestrian fatalities. The XGBoost model outperformed others, achieving a balanced accuracy of 98 %, accuracy of 90 %, precision of 92 %, recall of 90 %, and an F1 score of 91 %. Findings reveal that pedestrian fatalities are more common in mid-block locations and areas with poor visibility, with older adults and substance-impaired individuals at higher risk. These insights can inform policymakers and urban planners in implementing targeted safety measures, such as improved lighting, enhanced pedestrian infrastructure, and stricter traffic law enforcement, to reduce fatalities and improve public safety.
Authors: Weikai Wang, Erick Delage
Abstract: For continuing tasks, average cost Markov decision processes have well-documented value and can be solved using efficient algorithms. However, it explicitly assumes that the agent is risk-neutral. In this work, we extend risk-neutral algorithms to accommodate the more general class of dynamic risk measures. Specifically, we propose a relative value iteration (RVI) algorithm for planning and design two model-free Q-learning algorithms, namely a generic algorithm based on the multi-level Monte Carlo method, and an off-policy algorithm dedicated to utility-base shortfall risk measures. Both the RVI and MLMC-based Q-learning algorithms are proven to converge to optimality. Numerical experiments validate our analysis, confirms empirically the convergence of the off-policy algorithm, and demonstrate that our approach enables the identification of policies that are finely tuned to the intricate risk-awareness of the agent that they serve.
Authors: Bin Duan, Matthew B. Dwyer, Guowei Yang
Abstract: Deep Neural Networks (DNNs) have been widely employed across various domains, including safety-critical systems, necessitating comprehensive testing to ensure their reliability. Although numerous DNN model testing methods have been proposed to generate adversarial samples that are capable of revealing faults, existing methods typically perturb samples in the input space and then mutate these based on feedback from the DNN model. These methods often result in test samples that are not realistic and with low-probability reveal faults. To address these limitations, we propose a black-box DNN test input generation method, ARGUS, to generate realistic, diverse, and fault-revealing test inputs. ARGUS first compresses samples into a continuous latent space and then perturbs the original samples by interpolating these with samples of different classes. Subsequently, we employ a vector quantizer and decoder to reconstruct adversarial samples back into the input space. Additionally, we employ discriminators both in the latent space and in the input space to ensure the realism of the generated samples. Evaluation of ARGUS in comparison with state-of-the-art black-box testing and white-box testing methods, shows that ARGUS excels in generating realistic and diverse adversarial samples relative to the target dataset, and ARGUS successfully perturbs all original samples and achieves up to 4 times higher error rate than the best baseline method. Furthermore, using these adversarial samples for model retraining can improve model classification accuracy.
Authors: Jiali Cheng, Hadi Amiri
Abstract: Language models are prone to dataset biases, known as shortcuts and spurious correlations in data, which often result in performance drop on new data. We present a new debiasing framework called ``FairFlow'' that mitigates dataset biases by learning to be undecided in its predictions for data samples or representations associated with known or unknown biases. The framework introduces two key components: a suite of data and model perturbation operations that generate different biased views of input samples, and a contrastive objective that learns debiased and robust representations from the resulting biased views of samples. Experiments show that FairFlow outperforms existing debiasing methods, particularly against out-of-domain and hard test samples without compromising the in-domain performance
Authors: Mudit Gaur, Amrit Singh Bedi, Raghu Pasupathu, Vaneet Aggarwal
Abstract: Bilevel reinforcement learning (BRL) has emerged as a powerful mathematical framework for studying generative AI alignment and related problems. While several principled algorithmic frameworks have been proposed, key theoretical foundations, particularly those related to sample complexity, remain underexplored. Understanding and deriving tight sample complexity bounds are crucial for bridging the gap between theory and practice, guiding the development of more efficient algorithms. In this work, we present the first sample complexity result for BRL, achieving a bound of $\epsilon^{-4}$. This result extends to standard bilevel optimization problems, providing an interesting theoretical contribution with practical implications. To address the computational challenges associated with hypergradient estimation in bilevel optimization, we develop a first-order Hessian-free algorithm that does not rely on costly hypergradient computations. By leveraging matrix-free techniques and constrained optimization methods, our approach ensures scalability and practicality. Our findings pave the way for improved methods in AI alignment and other fields reliant on bilevel optimization.
Authors: Davide Villaboni, Alberto Castellini, Ivan Luciano Danesi, Alessandro Farinelli
Abstract: Transformer-based time series forecasting has recently gained strong interest due to the ability of transformers to model sequential data. Most of the state-of-the-art architectures exploit either temporal or inter-channel dependencies, limiting their effectiveness in multivariate time-series forecasting where both types of dependencies are crucial. We propose Sentinel, a full transformer-based architecture composed of an encoder able to extract contextual information from the channel dimension, and a decoder designed to capture causal relations and dependencies across the temporal dimension. Additionally, we introduce a multi-patch attention mechanism, which leverages the patching process to structure the input sequence in a way that can be naturally integrated into the transformer architecture, replacing the multi-head splitting process. Extensive experiments on standard benchmarks demonstrate that Sentinel, because of its ability to "monitor" both the temporal and the inter-channel dimension, achieves better or comparable performance with respect to state-of-the-art approaches.
Authors: Md. Shaheenur Islam Sumon, Md. Sakib Bin Islam, Md. Sohanur Rahman, Md. Sakib Abrar Hossain, Amith Khandakar, Anwarul Hasan, M Murugappan, Muhammad E. H. Chowdhury
Abstract: The early detection and prediction of cardiovascular diseases are crucial for reducing the severe morbidity and mortality associated with these conditions worldwide. A multi-headed self-attention mechanism, widely used in natural language processing (NLP), is operated by Transformers to understand feature interactions in feature spaces. However, the relationships between various features within biological systems remain ambiguous in these spaces, highlighting the necessity of early detection and prediction of cardiovascular diseases to reduce the severe morbidity and mortality with these conditions worldwide. We handle this issue with CardioTabNet, which exploits the strength of tab transformer to extract feature space which carries strong understanding of clinical cardiovascular data and its feature ranking. As a result, performance of downstream classical models significantly showed outstanding result. Our study utilizes the open-source dataset for heart disease prediction with 1190 instances and 11 features. In total, 11 features are divided into numerical (age, resting blood pressure, cholesterol, maximum heart rate, old peak, weight, and fasting blood sugar) and categorical (resting ECG, exercise angina, and ST slope). Tab transformer was used to extract important features and ranked them using random forest (RF) feature ranking algorithm. Ten machine-learning models were used to predict heart disease using selected features. After extracting high-quality features, the top downstream model (a hyper-tuned ExtraTree classifier) achieved an average accuracy rate of 94.1% and an average Area Under Curve (AUC) of 95.0%. Furthermore, a nomogram analysis was conducted to evaluate the model's effectiveness in cardiovascular risk assessment. A benchmarking study was conducted using state-of-the-art models to evaluate our transformer-driven framework.
Authors: Peijin Guo, Minghui Li, Hewen Pan, Ruixiang Huang, Lulu Xue, Shengqing Hu, Zikang Guo, Wei Wan, Shengshan Hu
Abstract: While deep learning models play a crucial role in predicting antibody-antigen interactions (AAI), the scarcity of publicly available sequence-structure pairings constrains their generalization. Current AAI methods often focus on residue-level static details, overlooking fine-grained structural representations of antibodies and their inter-antibody similarities. To tackle this challenge, we introduce a multi-modality representation approach that integates 3D structural and 1D sequence data to unravel intricate intra-antibody hierarchical relationships. By harnessing these representations, we present MuLAAIP, an AAI prediction framework that utilizes graph attention networks to illuminate graph-level structural features and normalized adaptive graph convolution networks to capture inter-antibody sequence associations. Furthermore, we have curated an AAI benchmark dataset comprising both structural and sequence information along with interaction labels. Through extensive experiments on this benchmark, our results demonstrate that MuLAAIP outperforms current state-of-the-art methods in terms of predictive performance. The implementation code and dataset are publicly available at https://github.com/trashTian/MuLAAIP for reproducibility.
Authors: Richa Rastogi, Yuta saito, Thorsten Joachims
Abstract: The feedback that AI systems (e.g., recommender systems, chatbots) collect from user interactions is a crucial source of training data. While short-term feedback (e.g., clicks, engagement) is widely used for training, there is ample evidence that optimizing short-term feedback does not necessarily achieve the desired long-term objectives. Unfortunately, directly optimizing for long-term objectives is challenging, and we identify the disconnect in the timescales of short-term interventions (e.g., rankings) and the long-term feedback (e.g., user retention) as one of the key obstacles. To overcome this disconnect, we introduce the framework of MultiScale Policy Learning to contextually reconcile that AI systems need to act and optimize feedback at multiple interdependent timescales. For any two levels, our formulation selects the shorter-term objective at the next lower scale to optimize the longer-term objective at the next higher scale. As a result, the policies at all levels effectively optimize for the long-term. We instantiate the framework with MultiScale Off-Policy Bandit Learning (MSBL) and demonstrate its effectiveness on three tasks relating to recommender systems and text generation.
Authors: Huitong Chen, Yu Wang, Yan Fan, Guosong Jiang, Qinghua Hu
Abstract: Class incremental learning (CIL) aims to enable models to continuously learn new classes without catastrophically forgetting old ones. A promising direction is to learn and use prototypes of classes during incremental updates. Despite simplicity and intuition, we find that such methods suffer from inadequate representation capability and unsatisfied feature overlap. These two factors cause class-wise confusion and limited performance. In this paper, we develop a Confusion-REduced AuTo-Encoder classifier (CREATE) for CIL. Specifically, our method employs a lightweight auto-encoder module to learn compact manifold for each class in the latent subspace, constraining samples to be well reconstructed only on the semantically correct auto-encoder. Thus, the representation stability and capability of class distributions are enhanced, alleviating the potential class-wise confusion problem. To further distinguish the overlapped features, we propose a confusion-aware latent space separation loss that ensures samples are closely distributed in their corresponding low-dimensional manifold while keeping away from the distributions of features from other classes. Our method demonstrates stronger representational capacity and discrimination ability by learning disentangled manifolds and reduces class confusion. Extensive experiments on multiple datasets and settings show that CREATE outperforms other state-of-the-art methods up to 5.41%.
Authors: Joshua E. Hammond, Tyler Soderstrom, Brian A. Korgel, Michael Baldea
Abstract: We present the Subset Extended Kalman Filter (SEKF) as a method to update previously trained model weights online rather than retraining or finetuning them when the system a model represents drifts away from the conditions under which it was trained. We identify the parameters to be updated using the gradient of the loss function and use the SEKF to update only these parameters. We compare finetuning and SEKF for online model maintenance in the presence of systemic drift through four dynamic regression case studies and find that the SEKF is able to maintain model accuracy as-well if not better than finetuning while requiring significantly less time per iteration, and less hyperparameter tuning.
Authors: Jiaming Ji, Xinyu Chen, Rui Pan, Han Zhu, Conghui Zhang, Jiahao Li, Donghai Hong, Boyuan Chen, Jiayi Zhou, Kaile Wang, Juntao Dai, Chi-Min Chan, Sirui Han, Yike Guo, Yaodong Yang
Abstract: Multimodal large language models (MLLMs) are critical for developing general-purpose AI assistants, yet they face growing safety risks. How can we ensure that MLLMs are safely aligned to prevent undesired behaviors such as discrimination, misinformation, or violations of ethical standards? In a further step, we need to explore how to fine-tune MLLMs to enhance reasoning performance while ensuring they satisfy safety constraints. Fundamentally, this can be formulated as a min-max optimization problem. In this study, we propose Safe RLHF-V, the first multimodal safety alignment framework that jointly optimizes helpfulness and safety using separate multimodal reward and cost models within a Lagrangian-based constrained optimization framework. Given that there is a lack of preference datasets that separate helpfulness and safety in multimodal scenarios, we introduce BeaverTails-V, the first open-source dataset with dual preference annotations for helpfulness and safety, along with multi-level safety labels (minor, moderate, severe). Additionally, we design a Multi-level Guardrail System to proactively defend against unsafe queries and adversarial attacks. By applying the Beaver-Guard-V moderation for 5 rounds of filtering and re-generation on the precursor model, the overall safety of the upstream model is significantly improved by an average of 40.9%. Experimental results demonstrate that fine-tuning different MLLMs with Safe RLHF can effectively enhance model helpfulness while ensuring improved safety. Specifically, Safe RLHF-V improves model safety by 34.2% and helpfulness by 34.3%. All of datasets, models, and code can be found at https://github.com/SafeRLHF-V to support the safety development of MLLMs and reduce potential societal risks.
Authors: Rebecca Clain, Eduardo Fernandes Montesuma, Fred Ngol\`e Mboula
Abstract: Decentralized Multi-Source Domain Adaptation (DMSDA) is a challenging task that aims to transfer knowledge from multiple related and heterogeneous source domains to an unlabeled target domain within a decentralized framework. Our work tackles DMSDA through a fully decentralized federated approach. In particular, we extend the Federated Dataset Dictionary Learning (FedDaDiL) framework by eliminating the necessity for a central server. FedDaDiL leverages Wasserstein barycenters to model the distributional shift across multiple clients, enabling effective adaptation while preserving data privacy. By decentralizing this framework, we enhance its robustness, scalability, and privacy, removing the risk of a single point of failure. We compare our method to its federated counterpart and other benchmark algorithms, showing that our approach effectively adapts source domains to an unlabeled target domain in a fully decentralized manner.
Authors: Abdullah Al-Khatib, Abdullah Ahmed, Klaus Moessner, Holger Timinger
Abstract: Onsite bandwidth reservation requests often face challenges such as price fluctuations and fairness issues due to unpredictable bandwidth availability and stringent latency requirements. Requesting bandwidth in advance can mitigate the impact of these fluctuations and ensure timely access to critical resources. In a multi-Mobile Network Operator (MNO) environment, vehicles need to select cost-effective and reliable resources for their safety-critical applications. This research aims to minimize resource costs by finding the best price among multiple MNOs. It formulates multi-operator scenarios as a Markov Decision Process (MDP), utilizing a Deep Reinforcement Learning (DRL) algorithm, specifically Dueling Deep Q-Learning. For efficient and stable learning, we propose a novel area-wise approach and an adaptive MDP synthetic close to the real environment. The Temporal Fusion Transformer (TFT) is used to handle time-dependent data and model training. Furthermore, the research leverages Amazon spot price data and adopts a multi-phase training approach, involving initial training on synthetic data, followed by real-world data. These phases enable the DRL agent to make informed decisions using insights from historical data and real-time observations. The results show that our model leads to significant cost reductions, up to 40%, compared to scenarios without a policy model in such a complex environment.
Authors: Triveni Magadum, Sanjana Murgod, Kartik Garg, Vivek Yadav, Harshit Mittal, Omkar Kushwaha
Abstract: In this research, renewable energy expansion in South America up to 2050 is predicted based on machine learning models that are trained on past energy data. The research employs gradient boosting regression and Prophet time series forecasting to make predictions of future generation capacities for solar, wind, hydroelectric, geothermal, biomass, and other renewable sources in South American nations. Model output analysis indicates staggering future expansion in the generation of renewable energy, with solar and wind energy registering the highest expansion rates. Geospatial visualization methods were applied to illustrate regional disparities in the utilization of renewable energy. The results forecast South America to record nearly 3-fold growth in the generation of renewable energy by the year 2050, with Brazil and Chile spearheading regional development. Such projections help design energy policy, investment strategy, and climate change mitigation throughout the region, in helping the developing economies to transition to sustainable energy.
Authors: Codefuse, Ling Team, :, Wenting Cai, Yuchen Cao, Chaoyu Chen, Chen Chen, Siba Chen, Qing Cui, Peng Di, Junpeng Fang, Zi Gong, Ting Guo, Zhengyu He, Yang Huang, Cong Li, Jianguo Li, Zheng Li, Shijie Lian, BingChang Liu, Songshan Luo, Shuo Mao, Min Shen, Jian Wu, Jiaolong Yang, Wenjie Yang, Tong Ye, Hang Yu, Wei Zhang, Zhenduo Zhang, Hailin Zhao, Xunjin Zheng, Jun Zhou
Abstract: Recent advancements in code large language models (LLMs) have demonstrated remarkable capabilities in code generation and understanding. It is still challenging to build a code LLM with comprehensive performance yet ultimate efficiency. Many attempts have been released in the open source community to break the trade-off between performance and efficiency, such as the Qwen Coder series and the DeepSeek Coder series. This paper introduces yet another attempt in this area, namely Ling-Coder-Lite. We leverage the efficient Mixture-of-Experts (MoE) architecture along with a set of high-quality data curation methods (especially those based on program analytics) to build an efficient yet powerful code LLM. Ling-Coder-Lite exhibits on-par performance on 12 representative coding benchmarks compared to state-of-the-art models of similar size, such as Qwen2.5-Coder-7B and DeepSeek-Coder-V2-Lite, while offering competitive latency and throughput. In practice, we achieve a 50\% reduction in deployment resources compared to the similar-sized dense model without performance loss. To facilitate further research and development in this area, we open-source our models as well as a substantial portion of high-quality data for the annealing and post-training stages. The models and data can be accessed at~\url{https://huggingface.co/inclusionAI/Ling-Coder-lite}.
Authors: Chaoyu Liu, Davide Murari, Chris Budd, Lihao Liu, Carola-Bibiane Sch\"onlieb
Abstract: Partial Differential Equation (PDE) problems often exhibit strong local spatial structures, and effectively capturing these structures is critical for approximating their solutions. Recently, the Fourier Neural Operator (FNO) has emerged as an efficient approach for solving these PDE problems. By using parametrization in the frequency domain, FNOs can efficiently capture global patterns. However, this approach inherently overlooks the critical role of local spatial features, as frequency-domain parameterized convolutions primarily emphasize global interactions without encoding comprehensive localized spatial dependencies. Although several studies have attempted to address this limitation, their extracted Local Spatial Features (LSFs) remain insufficient, and computational efficiency is often compromised. To address this limitation, we introduce a convolutional neural network (CNN) preprocessor to extract LSFs directly from input data, resulting in a hybrid architecture termed \textit{Conv-FNO}. Furthermore, we introduce two novel resizing schemes to make our Conv-FNO resolution invariant. In this work, we focus on demonstrating the effectiveness of incorporating LSFs into FNOs by conducting both a theoretical analysis and extensive numerical experiments. Our findings show that this simple yet impactful modification enhances the representational capacity of FNOs and significantly improves performance on challenging PDE benchmarks.
Authors: Giovanni Briglia, Stefano Mariani, Franco Zambonelli
Abstract: Causal reasoning is increasingly used in Reinforcement Learning (RL) to improve the learning process in several dimensions: efficacy of learned policies, efficiency of convergence, generalisation capabilities, safety and interpretability of behaviour. However, applications of causal reasoning to Multi-Agent RL (MARL) are still mostly unexplored. In this paper, we take the first step in investigating the opportunities and challenges of applying causal reasoning in MARL. We measure the impact of a simple form of causal augmentation in state-of-the-art MARL scenarios increasingly requiring cooperation, and with state-of-the-art MARL algorithms exploiting various degrees of collaboration between agents. Then, we discuss the positive as well as negative results achieved, giving us the chance to outline the areas where further research may help to successfully transfer causal RL to the multi-agent setting.
Authors: Z. Zarezadeh, N. Zarezadeh
Abstract: In this paper we consider a new probability sampling methods based on Langevin diffusion dynamics to resolve the problem of existing Monte Carlo algorithms when draw samples from high dimensional target densities. We extent Metropolis-Adjusted Langevin Diffusion algorithm by modelling the stochasticity of precondition matrix as a random matrix. An advantage compared to other proposal method is that it only requires the gradient of log-posterior. The proposed method provides fully adaptation mechanisms to tune proposal densities to exploits and adapts the geometry of local structures of statistical models. We clarify the benefits of the new proposal by modelling a Quantum Probability Density Functions of a free particle in a plane (energy Eigen-functions). The proposed model represents a remarkable improvement in terms of performance accuracy and computational time over standard MCMC method.
Authors: Zeyu Jia, Yury Polyanskiy, Alexander Rakhlin
Abstract: We study the problem of sequential probability assignment under logarithmic loss, both with and without side information. Our objective is to analyze the minimax regret -- a notion extensively studied in the literature -- in terms of geometric quantities, such as covering numbers and scale-sensitive dimensions. We show that the minimax regret for the case of no side information (equivalently, the Shtarkov sum) can be upper bounded in terms of sequential square-root entropy, a notion closely related to Hellinger distance. For the problem of sequential probability assignment with side information, we develop both upper and lower bounds based on the aforementioned entropy. The lower bound matches the upper bound, up to log factors, for classes in the Donsker regime (according to our definition of entropy).
Authors: Maryam Abdolali, Romina Zakerian, Behnam Roshanfekr, Fardin Ayar, Mohammad Rahmati
Abstract: In this paper, we propose a novel framework that combines ensemble learning with augmented graph structures to improve the performance and robustness of semi-supervised node classification in graphs. By creating multiple augmented views of the same graph, our approach harnesses the "wisdom of a diverse crowd", mitigating the challenges posed by noisy graph structures. Leveraging ensemble learning allows us to simultaneously achieve three key goals: adaptive confidence threshold selection based on model agreement, dynamic determination of the number of high-confidence samples for training, and robust extraction of pseudo-labels to mitigate confirmation bias. Our approach uniquely integrates adaptive ensemble consensus to flexibly guide pseudo-label extraction and sample selection, reducing the risks of error accumulation and improving robustness. Furthermore, the use of ensemble-driven consensus for pseudo-labeling captures subtle patterns that individual models often overlook, enabling the model to generalize better. Experiments on several real-world datasets demonstrate the effectiveness of our proposed method.
Authors: Andrei V. Konstantinov, Lev V. Utkin
Abstract: There are many approaches for training decision trees. This work introduces a novel gradient-based method for constructing decision trees that optimize arbitrary differentiable loss functions, overcoming the limitations of heuristic splitting rules. Unlike traditional approaches that rely on heuristic splitting rules, the proposed method refines predictions using the first and second derivatives of the loss function, enabling the optimization of complex tasks such as classification, regression, and survival analysis. We demonstrate the method's applicability to classification, regression, and survival analysis tasks, including those with censored data. Numerical experiments on both real and synthetic datasets compare the proposed method with traditional decision tree algorithms, such as CART, Extremely Randomized Trees, and SurvTree. The implementation of the method is publicly available, providing a practical tool for researchers and practitioners. This work advances the field of decision tree-based modeling, offering a more flexible and accurate approach for handling structured data and complex tasks. By leveraging gradient-based optimization, the proposed method bridges the gap between traditional decision trees and modern machine learning techniques, paving the way for further innovations in interpretable and high-performing models.
Authors: Yali Fu, Jindong Li, Qi Wang, Qianli Xing
Abstract: Unsupervised graph-level anomaly detection (UGLAD) is a critical and challenging task across various domains, such as social network analysis, anti-cancer drug discovery, and toxic molecule identification. However, existing methods often struggle to capture the long-range dependencies efficiently and neglect the spectral information. Recently, selective State Space Models (SSMs), particularly Mamba, have demonstrated remarkable advantages in capturing long-range dependencies with linear complexity and a selection mechanism. Motivated by their success across various domains, we propose GLADMamba, a novel framework that adapts the selective state space model into UGLAD field. We design View-Fused Mamba (VFM) with a Mamba-Transformer-style architecture to efficiently fuse information from different views with a selective state mechanism. We also design Spectrum-Guided Mamba (SGM) with a Mamba-Transformer-style architecture to leverage the Rayleigh quotient to guide the embedding refining process. GLADMamba can dynamically focus on anomaly-related information while discarding irrelevant information for anomaly detection. To the best of our knowledge, this is the first work to introduce Mamba and explicit spectral information to UGLAD. Extensive experiments on 12 real-world datasets demonstrate that GLADMamba outperforms existing state-of-the-art methods, achieving superior performance in UGLAD. The code is available at https://github.com/Yali-F/GLADMamba.
Authors: Luke McDermott, Rahul Parhi
Abstract: Recent works have shown that Dataset Distillation, the process for summarizing the training data, can be leveraged to accelerate the training of deep learning models. However, its impact on training dynamics, particularly in neural network pruning, remains largely unexplored. In our work, we use distilled data in the inner loop of iterative magnitude pruning to produce sparse, trainable subnetworks at initialization -- more commonly known as lottery tickets. While using 150x less training points, our algorithm matches the performance of traditional lottery ticket rewinding on ResNet-18 & CIFAR-10. Previous work highlights that lottery tickets can be found when the dense initialization is stable to SGD noise (i.e. training across different ordering of the data converges to the same minima). We extend this discovery, demonstrating that stable subnetworks can exist even within an unstable dense initialization. In our linear mode connectivity studies, we find that pruning with distilled data discards parameters that contribute to the sharpness of the loss landscape. Lastly, we show that by first generating a stable sparsity mask at initialization, we can find lottery tickets at significantly higher sparsities than traditional iterative magnitude pruning.
Authors: Yongqi Huang, Jitao Zhao, Dongxiao He, Di Jin, Yuxiao Huang, Zhen Wang
Abstract: Graph Contrastive Learning (GCL) aims to self-supervised learn low-dimensional graph representations, primarily through instance discrimination, which involves manually mining positive and negative pairs from graphs, increasing the similarity of positive pairs while decreasing negative pairs. Drawing from the success of Contrastive Learning (CL) in other domains, a consensus has been reached that the effectiveness of GCLs depends on a large number of negative pairs. As a result, despite the significant computational overhead, GCLs typically leverage as many negative node pairs as possible to improve model performance. However, given that nodes within a graph are interconnected, we argue that nodes cannot be treated as independent instances. Therefore, we challenge this consensus: Does employing more negative nodes lead to a more effective GCL model? To answer this, we explore the role of negative nodes in the commonly used InfoNCE loss for GCL and observe that: (1) Counterintuitively, a large number of negative nodes can actually hinder the model's ability to distinguish nodes with different semantics. (2) A smaller number of high-quality and non-topologically coupled negative nodes are sufficient to enhance the discriminability of representations. Based on these findings, we propose a new method called GCL with Effective and Efficient Negative samples, E2Neg, which learns discriminative representations using only a very small set of representative negative samples. E2Neg significantly reduces computational overhead and speeds up model training. We demonstrate the effectiveness and efficiency of E2Neg across multiple datasets compared to other GCL methods.
Authors: Koustubh Phalak, Junde Li, Swaroop Ghosh
Abstract: Training Quantum Neural Networks (QNNs) on large amount of classical data can be both time consuming as well as expensive. Higher amount of training data would require higher number of gradient descent steps to reach convergence. This, in turn would imply that the QNN will require higher number of quantum executions, thereby driving up its overall execution cost. In this work, we propose performing the dataset distillation process for QNNs, where we use a novel quantum variant of classical LeNet model containing residual connection and trainable Hermitian observable in the Parametric Quantum Circuit (PQC) of the QNN. This approach yields highly informative yet small number of training data at similar performance as the original data. We perform distillation for MNIST and Cifar-10 datasets, and on comparison with classical models observe that both the datasets yield reasonably similar post-inferencing accuracy on quantum LeNet (91.9% MNIST, 50.3% Cifar-10) compared to classical LeNet (94% MNIST, 54% Cifar-10). We also introduce a non-trainable Hermitian for ensuring stability in the distillation process and note marginal reduction of up to 1.8% (1.3%) for MNIST (Cifar-10) dataset.
Authors: Sami Zhioua, Ruta Binkyte, Ayoub Ouni, Farah Barika Ktata
Abstract: Accurately measuring discrimination is crucial to faithfully assessing fairness of trained machine learning (ML) models. Any bias in measuring discrimination leads to either amplification or underestimation of the existing disparity. Several sources of bias exist and it is assumed that bias resulting from machine learning is born equally by different groups (e.g. females vs males, whites vs blacks, etc.). If, however, bias is born differently by different groups, it may exacerbate discrimination against specific sub-populations. Sampling bias, in particular, is inconsistently used in the literature to describe bias due to the sampling procedure. In this paper, we attempt to disambiguate this term by introducing clearly defined variants of sampling bias, namely, sample size bias (SSB) and underrepresentation bias (URB). Through an extensive set of experiments on benchmark datasets and using mainstream learning algorithms, we expose relevant observations in several model training scenarios. The observations are finally framed as actionable recommendations for practitioners.
Authors: Alexandre Perez-Lebel, Gael Varoquaux, Sanmi Koyejo, Matthieu Doutreligne, Marine Le Morvan
Abstract: Probabilistic classifiers are central for making informed decisions under uncertainty. Based on the maximum expected utility principle, optimal decision rules can be derived using the posterior class probabilities and misclassification costs. Yet, in practice only learned approximations of the oracle posterior probabilities are available. In this work, we quantify the excess risk (a.k.a. regret) incurred using approximate posterior probabilities in batch binary decision-making. We provide analytical expressions for miscalibration-induced regret ($R^{\mathrm{CL}}$), as well as tight and informative upper and lower bounds on the regret of calibrated classifiers ($R^{\mathrm{GL}}$). These expressions allow us to identify regimes where recalibration alone addresses most of the regret, and regimes where the regret is dominated by the grouping loss, which calls for post-training beyond recalibration. Crucially, both $R^{\mathrm{CL}}$ and $R^{\mathrm{GL}}$ can be estimated in practice using a calibration curve and a recent grouping loss estimator. On NLP experiments, we show that these quantities identify when the expected gain of more advanced post-training is worth the operational cost. Finally, we highlight the potential of multicalibration approaches as efficient alternatives to costlier fine-tuning approaches.
Authors: Xiaochen Zhang, Haoyi Xiong
Abstract: In high-dimensional and high-stakes contexts, ensuring both rigorous statistical guarantees and interpretability in feature extraction from complex tabular data remains a formidable challenge. Traditional methods such as Principal Component Analysis (PCA) reduce dimensionality and identify key features that explain the most variance, but are constrained by their reliance on linear assumptions. In contrast, neural networks offer assumption-free feature extraction through self-supervised learning techniques such as autoencoders, though their interpretability remains a challenge in fields requiring transparency. To address this gap, this paper introduces Spofe, a novel self-supervised machine learning pipeline that marries the power of kernel principal components for capturing nonlinear dependencies with a sparse and principled polynomial representation to achieve clear interpretability with statistical rigor. Underpinning our approach is a robust theoretical framework that delivers precise error bounds and rigorous false discovery rate (FDR) control via a multi-objective knockoff selection procedure; it effectively bridges the gap between data-driven complexity and statistical reliability via three stages: (1) generating self-supervised signals using kernel principal components to model complex patterns, (2) distilling these signals into sparse polynomial functions for improved interpretability, and (3) applying a multi-objective knockoff selection procedure with significance testing to rigorously identify important features. Extensive experiments on diverse real-world datasets demonstrate the effectiveness of Spofe, consistently surpassing KPCA, SKPCA, and other methods in feature selection for regression and classification tasks. Visualization and case studies highlight its ability to uncover key insights, enhancing interpretability and practical utility.
Authors: Zeyuan Ma, Zhiyang Huang, Jiacheng Chen, Zhiguang Cao, Yue-Jiao Gong
Abstract: Recent Meta-Black-Box Optimization (MetaBBO) approaches have shown possibility of enhancing the optimization performance through learning meta-level policies to dynamically configure low-level optimizers. However, existing MetaBBO approaches potentially consume massive function evaluations to train their meta-level policies. Inspired by the recent trend of using surrogate models for cost-friendly evaluation of expensive optimization problems, in this paper, we propose a novel MetaBBO framework which combines surrogate learning process and reinforcement learning-aided Differential Evolution algorithm, namely Surr-RLDE, to address the intensive function evaluation in MetaBBO. Surr-RLDE comprises two learning stages: surrogate learning and policy learning. In surrogate learning, we train a Kolmogorov-Arnold Networks (KAN) with a novel relative-order-aware loss to accurately approximate the objective functions of the problem instances used for subsequent policy learning. In policy learning, we employ reinforcement learning (RL) to dynamically configure the mutation operator in DE. The learned surrogate model is integrated into the training of the RL-based policy to substitute for the original objective function, which effectively reduces consumed evaluations during policy learning. Extensive benchmark results demonstrate that Surr-RLDE not only shows competitive performance to recent baselines, but also shows compelling generalization for higher-dimensional problems. Further ablation studies underscore the effectiveness of each technical components in Surr-RLDE. We open-source Surr-RLDE at https://github.com/GMC-DRL/Surr-RLDE.
Authors: Xiaoming Qi, Jingyang Zhang, Huazhu Fu, Guanyu Yang, Shuo Li, Yueming Jin
Abstract: Federated continual learning (FCL) offers an emerging pattern to facilitate the applicability of federated learning (FL) in real-world scenarios, where tasks evolve dynamically and asynchronously across clients, especially in medical scenario. Existing server-side FCL methods in nature domain construct a continually learnable server model by client aggregation on all-involved tasks. However, they are challenged by: (1) Catastrophic forgetting for previously learned tasks, leading to error accumulation in server model, making it difficult to sustain comprehensive knowledge across all tasks. (2) Biased optimization due to asynchronous tasks handled across different clients, leading to the collision of optimization targets of different clients at the same time steps. In this work, we take the first step to propose a novel server-side FCL pattern in medical domain, Dynamic Allocation Hypernetwork with adaptive model recalibration (\textbf{FedDAH}). It is to facilitate collaborative learning under the distinct and dynamic task streams across clients. To alleviate the catastrophic forgetting, we propose a dynamic allocation hypernetwork (DAHyper) where a continually updated hypernetwork is designed to manage the mapping between task identities and their associated model parameters, enabling the dynamic allocation of the model across clients. For the biased optimization, we introduce a novel adaptive model recalibration (AMR) to incorporate the candidate changes of historical models into current server updates, and assign weights to identical tasks across different time steps based on the similarity for continual optimization. Extensive experiments on the AMOS dataset demonstrate the superiority of our FedDAH to other FCL methods on sites with different task streams. The code is available:https://github.com/jinlab-imvr/FedDAH.
Authors: Shahaf Bassan, Shlomit Gur, Sergey Zeltyn, Konstantinos Mavrogiorgos, Ron Eliav, Dimosthenis Kyriazis
Abstract: Tasks in Predictive Business Process Monitoring (PBPM), such as Next Activity Prediction, focus on generating useful business predictions from historical case logs. Recently, Deep Learning methods, particularly sequence-to-sequence models like Long Short-Term Memory (LSTM), have become a dominant approach for tackling these tasks. However, to enhance model transparency, build trust in the predictions, and gain a deeper understanding of business processes, it is crucial to explain the decisions made by these models. Existing explainability methods for PBPM decisions are typically *post-hoc*, meaning they provide explanations only after the model has been trained. Unfortunately, these post-hoc approaches have shown to face various challenges, including lack of faithfulness, high computational costs and a significant sensitivity to out-of-distribution samples. In this work, we introduce, to the best of our knowledge, the first *self-explaining neural network* architecture for predictive process monitoring. Our framework trains an LSTM model that not only provides predictions but also outputs a concise explanation for each prediction, while adapting the optimization objective to improve the reliability of the explanation. We first demonstrate that incorporating explainability into the training process does not hurt model performance, and in some cases, actually improves it. Additionally, we show that our method outperforms post-hoc approaches in terms of both the faithfulness of the generated explanations and substantial improvements in efficiency.
Authors: Massimiliano Ghiotto
Abstract: This paper introduces HyperNOs, a PyTorch library designed to streamline and automate the process of exploring neural operators, with a special focus on hyperparameter optimization for comprehensive and exhaustive exploration. Indeed, HyperNOs takes advantage of state-of-the-art optimization algorithms and parallel computing implemented in the Ray-tune library to efficiently explore the hyperparameter space of neural operators. We also implement many useful functionalities for studying neural operators with a user-friendly interface, such as the possibility to train the model with a fixed number of parameters or to train the model with multiple datasets and different resolutions. We integrate Fourier neural operators and convolutional neural operators in our library, achieving state of the art results on many representative benchmarks, demonstrating the capabilities of HyperNOs to handle real datasets and modern architectures. The library is designed to be easy to use with the provided model and datasets, but also to be easily extended to use new datasets and custom neural operator architectures.
Authors: Chi-Ning Chou, Hang Le, Yichen Wang, SueYeon Chung
Abstract: The ability to integrate task-relevant information into neural representations is a fundamental aspect of both biological and artificial intelligence. To enable theoretical analysis, recent work has examined whether a network learns task-relevant features (rich learning) or resembles a random feature model (or a kernel machine, i.e., lazy learning). However, this simple lazy-versus-rich dichotomy overlooks the possibility of various subtypes of feature learning that emerge from different architectures, learning rules, and data properties. Furthermore, most existing approaches emphasize weight matrices or neural tangent kernels, limiting their applicability to neuroscience because they do not explicitly characterize representations. In this work, we introduce an analysis framework based on representational geometry to study feature learning. Instead of analyzing what are the learned features, we focus on characterizing how task-relevant representational manifolds evolve during the learning process. In both theory and experiment, we find that when a network learns features useful for solving a task, the task-relevant manifolds become increasingly untangled. Moreover, by tracking changes in the underlying manifold geometry, we uncover distinct learning stages throughout training, as well as different learning strategies associated with training hyperparameters, uncovering subtypes of feature learning beyond the lazy-versus-rich dichotomy. Applying our method to neuroscience and machine learning, we gain geometric insights into the structural inductive biases of neural circuits solving cognitive tasks and the mechanisms underlying out-of-distribution generalization in image classification. Our framework provides a novel geometric perspective for understanding and quantifying feature learning in both artificial and biological neural networks.
Authors: Juntao Dai, Taiye Chen, Yaodong Yang, Qian Zheng, Gang Pan
Abstract: Reinforcement learning from human feedback (RLHF) is an effective method for aligning large language models (LLMs) with human values. However, reward over-optimization remains an open challenge leading to discrepancies between the performance of LLMs under the reward model and the true human objectives. A primary contributor to reward over-optimization is the extrapolation error that arises when the reward model evaluates out-of-distribution (OOD) responses. However, current methods still fail to prevent the increasing frequency of OOD response generation during the reinforcement learning (RL) process and are not effective at handling extrapolation errors from OOD responses. In this work, we propose the Behavior-Supported Policy Optimization (BSPO) method to mitigate the reward over-optimization issue. Specifically, we define behavior policy as the next token distribution of the reward training dataset to model the in-distribution (ID) region of the reward model. Building on this, we introduce the behavior-supported Bellman operator to regularize the value function, penalizing all OOD values without impacting the ID ones. Consequently, BSPO reduces the generation of OOD responses during the RL process, thereby avoiding overestimation caused by the reward model's extrapolation errors. Theoretically, we prove that BSPO guarantees a monotonic improvement of the supported policy until convergence to the optimal behavior-supported policy. Empirical results from extensive experiments show that BSPO outperforms baselines in preventing reward over-optimization due to OOD evaluation and finding the optimal ID policy.
Authors: Xiaojie Yang, Zipei Fan, Hangli Ge, Takashi Michikata, Ryosuke Shibasaki, Noboru Koshizuka
Abstract: Human mobility data are fused with multiple travel patterns and hidden spatiotemporal patterns are extracted by integrating user, location, and time information to improve next location prediction accuracy. In existing next location prediction methods, different causal relationships that result from patterns in human mobility data are ignored, which leads to confounding information that can have a negative effect on predictions. Therefore, this study introduces a causality-aware framework for next location prediction, focusing on human mobility stratification for travel patterns. In our research, a novel causal graph is developed that describes the relationships between various input variables. We use counterfactuals to enhance the indirect effects in our causal graph for specific travel patterns: non-anchor targeted travels. The proposed framework is designed as a plug-and-play module that integrates multiple next location prediction paradigms. We tested our proposed framework using several state-of-the-art models and human mobility datasets, and the results reveal that the proposed module improves the prediction performance. In addition, we provide results from the ablation study and quantitative study to demonstrate the soundness of our causal graph and its ability to further enhance the interpretability of the current next location prediction models.
Authors: Edgar Torres, Jonathan Schiefer, Mathias Niepert
Abstract: Physics-informed neural networks (PINNs) have emerged as a promising approach to solving partial differential equations (PDEs) using neural networks, particularly in data-scarce scenarios, due to their unsupervised training capability. However, limitations related to convergence and the need for re-optimization with each change in PDE parameters hinder their widespread adoption across scientific and engineering applications. This survey reviews existing research that addresses these limitations through transfer learning and meta-learning. The covered methods improve the training efficiency, allowing faster adaptation to new PDEs with fewer data and computational resources. While traditional numerical methods solve systems of differential equations directly, neural networks learn solutions implicitly by adjusting their parameters. One notable advantage of neural networks is their ability to abstract away from specific problem domains, allowing them to retain, discard, or adapt learned representations to efficiently address similar problems. By exploring the application of these techniques to PINNs, this survey identifies promising directions for future research to facilitate the broader adoption of PINNs in a wide range of scientific and engineering applications.
Authors: Hongliang Chi, Qiong Wu, Zhengyi Zhou, Yao Ma
Abstract: Graph Neural Networks (GNNs) have demonstrated remarkable performance in various graph-based machine learning tasks, yet evaluating the importance of neighbors of testing nodes remains largely unexplored due to the challenge of assessing data importance without test labels. To address this gap, we propose Shapley-Guided Utility Learning (SGUL), a novel framework for graph inference data valuation. SGUL innovatively combines transferable data-specific and modelspecific features to approximate test accuracy without relying on ground truth labels. By incorporating Shapley values as a preprocessing step and using feature Shapley values as input, our method enables direct optimization of Shapley value prediction while reducing computational demands. SGUL overcomes key limitations of existing methods, including poor generalization to unseen test-time structures and indirect optimization. Experiments on diverse graph datasets demonstrate that SGUL consistently outperforms existing baselines in both inductive and transductive settings. SGUL offers an effective, efficient, and interpretable approach for quantifying the value of test-time neighbors.
Authors: Ziheng Chen, Jiali Cheng, Gabriele Tolomei, Sijia Liu, Hadi Amiri, Yu Wang, Kaushiki Nag, Lu Lin
Abstract: As compliance with privacy regulations becomes increasingly critical, the growing demand for data privacy has highlighted the significance of machine unlearning in many real world applications, such as social network and recommender systems, many of which can be represented as graph-structured data. However, existing graph unlearning algorithms indiscriminately modify edges or nodes from well-trained models without considering the potential impact of such structural modifications on fairness. For example, forgetting links between nodes with different genders in a social network may exacerbate group disparities, leading to significant fairness concerns. To address these challenges, we propose a novel approach that jointly optimizes the graph structure and the corresponding model for fair unlearning tasks. Specifically,our approach rewires the graph to enhance unlearning efficiency by removing redundant edges that hinder forgetting while preserving fairness through targeted edge augmentation. Additionally, we introduce a worst-case evaluation mechanism to assess the reliability of fair unlearning performance. Extensive experiments on real-world datasets demonstrate the effectiveness of the proposed approach in achieving superior unlearning outcomes.
Authors: Georg Ziegner, Michael Choi, Hung Mac Chan Le, Sahil Sakhuja, Arash Sarmadi
Abstract: Multi-echelon inventory optimization (MEIO) is critical for effective supply chain management, but its inherent complexity can pose significant challenges. Heuristics are commonly used to address this complexity, yet they often face limitations in scope and scalability. Recent research has found deep reinforcement learning (DRL) to be a promising alternative to traditional heuristics, offering greater versatility by utilizing dynamic decision-making capabilities. However, since DRL is known to struggle with the curse of dimensionality, its relevance to complex real-life supply chain scenarios is still to be determined. This thesis investigates DRL's applicability to MEIO problems of increasing complexity. A state-of-the-art DRL model was replicated, enhanced, and tested across 13 supply chain scenarios, combining diverse network structures and parameters. To address DRL's challenges with dimensionality, additional models leveraging graph neural networks (GNNs) and multi-agent reinforcement learning (MARL) were developed, culminating in the novel iterative multi-agent reinforcement learning (IMARL) approach. IMARL demonstrated superior scalability, effectiveness, and reliability in optimizing inventory policies, consistently outperforming benchmarks. These findings confirm the potential of DRL, particularly IMARL, to address real-world supply chain challenges and call for additional research to further expand its applicability.
Authors: Nitish Dashora, Dibya Ghosh, Sergey Levine
Abstract: Online reinforcement learning (RL) with sparse rewards poses a challenge partly because of the lack of feedback on states leading to the goal. Furthermore, expert offline data with reward signal is rarely available to provide this feedback and bootstrap online learning. How can we guide online agents to the right solution without this on-task data? Reward shaping offers a solution by providing fine-grained signal to nudge the policy towards the optimal solution. However, reward shaping often requires domain knowledge to hand-engineer heuristics for a specific goal. To enable more general and inexpensive guidance, we propose and analyze a data-driven methodology that automatically guides RL by learning from widely available video data such as Internet recordings, off-task demonstrations, task failures, and undirected environment interaction. By learning a model of optimal goal-conditioned value from diverse passive data, we open the floor to scaling up and using various data sources to model general goal-reaching behaviors relevant to guiding online RL. Specifically, we use intent-conditioned value functions to learn from diverse videos and incorporate these goal-conditioned values into the reward. Our experiments show that video-trained value functions work well with a variety of data sources, exhibit positive transfer from human video pre-training, can generalize to unseen goals, and scale with dataset size.
Authors: Roberto Garcia, Jerry Liu, Daniel Sorvisto, Sabri Eyuboglu
Abstract: Large Language Models (LLMs) are computationally intensive, particularly during inference. Neuron-adaptive techniques, which selectively activate neurons in Multi-Layer Perceptron (MLP) layers, offer some speedups but suffer from limitations in modern Transformers. These include reliance on sparse activations, incompatibility with attention layers, and the use of costly neuron masking techniques. To address these issues, we propose the Adaptive Rank Allocation framework and introduce the Rank and Neuron Allocator (RaNA) adapter. RaNA adapters leverage rank adapters, which operate on linear layers by applying both low-rank matrix decompositions and adaptive masking to efficiently allocate compute without depending on activation sparsity. This enables RaNA to be generally applied to MLPs and linear components of attention modules, while eliminating the need for expensive maskers found in neuron-adaptive methods. Notably, when compared to neuron adapters, RaNA improves perplexity by up to 7 points and increases accuracy by up to 8 percentage-points when reducing FLOPs by $\sim$44% in state-of-the-art Transformer architectures. These results position RaNA as a robust solution for improving inference efficiency in modern Transformer architectures.
Authors: Philipp Grohs, Samuel Lanthaler, Margaret Trautner
Abstract: This work studies the sampling complexity of learning with ReLU neural networks and neural operators. For mappings belonging to relevant approximation spaces, we derive upper bounds on the best-possible convergence rate of any learning algorithm, with respect to the number of samples. In the finite-dimensional case, these bounds imply a gap between the parametric and sampling complexities of learning, known as the \emph{theory-to-practice gap}. In this work, a unified treatment of the theory-to-practice gap is achieved in a general $L^p$-setting, while at the same time improving available bounds in the literature. Furthermore, based on these results the theory-to-practice gap is extended to the infinite-dimensional setting of operator learning. Our results apply to Deep Operator Networks and integral kernel-based neural operators, including the Fourier neural operator. We show that the best-possible convergence rate in a Bochner $L^p$-norm is bounded by Monte-Carlo rates of order $1/p$.
Authors: Shubhankar Agarwal, Hamzah I. Khan, Sandeep P. Chinchali, David Fridovich-Keil
Abstract: Saddle point optimization is a critical problem employed in numerous real-world applications, including portfolio optimization, generative adversarial networks, and robotics. It has been extensively studied in cases where the objective function is known and differentiable. Existing work in black-box settings with unknown objectives that can only be sampled either assumes convexity-concavity in the objective to simplify the problem or operates with noisy gradient estimators. In contrast, we introduce a framework inspired by Bayesian optimization which utilizes Gaussian processes to model the unknown (potentially nonconvex-nonconcave) objective and requires only zeroth-order samples. Our approach frames the saddle point optimization problem as a two-level process which can flexibly integrate existing and novel approaches to this problem. The upper level of our framework produces a model of the objective function by sampling in promising locations, and the lower level of our framework uses the existing model to frame and solve a general-sum game to identify locations to sample. This lower level procedure can be designed in complementary ways, and we demonstrate the flexibility of our approach by introducing variants which appropriately trade off between factors like runtime, the cost of function evaluations, and the number of available initial samples. We experimentally demonstrate these algorithms on synthetic and realistic datasets in black-box nonconvex-nonconcave settings, showcasing their ability to efficiently locate local saddle points in these contexts.
Authors: Massimo Bini, Leander Girrbach, Zeynep Akata
Abstract: Parameter-Efficient FineTuning (PEFT) methods have recently gained significant popularity thanks to the widespread availability of large-scale pretrained models. These methods allow for quick adaptation to downstream tasks with minimal computational cost. However, popular finetuning methods such as LoRA exhibit limited robustness when it comes to hyperparameter choices or extended training regimes, preventing optimal out-of-the-box performance. In contrast, bounded approaches, such as ETHER, provide greater robustness but are limited to extremely low-rank adaptations and fixed-strength transformations, reducing their adaptation expressive power. In this work, we propose Decoupled Low-rank Adaptation (DeLoRA), a novel finetuning method that normalizes and scales learnable low-rank matrices. By bounding the distance of the transformation, DeLoRA effectively decouples the angular learning from the adaptation strength, enhancing robustness without compromising performance. Through evaluations on subject-driven image generation, natural language understanding, and instruction tuning, we show that DeLoRA matches or surpasses performance of competing PEFT methods, while exhibiting stronger robustness. Code is available at https://github.com/ExplainableML/DeLoRA.
Authors: Akash Agrawal (Carnegie Mellon University), Christopher McComb (Carnegie Mellon University)
Abstract: Multi-fidelity Reinforcement Learning (RL) frameworks efficiently utilize computational resources by integrating analysis models of varying accuracy and costs. The prevailing methodologies, characterized by transfer learning, human-inspired strategies, control variate techniques, and adaptive sampling, predominantly depend on a structured hierarchy of models. However, this reliance on a model hierarchy can exacerbate variance in policy learning when the underlying models exhibit heterogeneous error distributions across the design space. To address this challenge, this work proposes a novel adaptive multi-fidelity RL framework, in which multiple heterogeneous, non-hierarchical low-fidelity models are dynamically leveraged alongside a high-fidelity model to efficiently learn a high-fidelity policy. Specifically, low-fidelity policies and their experience data are adaptively used for efficient targeted learning, guided by their alignment with the high-fidelity policy. The effectiveness of the approach is demonstrated in an octocopter design optimization problem, utilizing two low-fidelity models alongside a high-fidelity simulator. The results demonstrate that the proposed approach substantially reduces variance in policy learning, leading to improved convergence and consistent high-quality solutions relative to traditional hierarchical multi-fidelity RL methods. Moreover, the framework eliminates the need for manually tuning model usage schedules, which can otherwise introduce significant computational overhead. This positions the framework as an effective variance-reduction strategy for multi-fidelity RL, while also mitigating the computational and operational burden of manual fidelity scheduling.
Authors: Shih-Min Yang, Martin Magnusson, Johannes A. Stork, Todor Stoyanov
Abstract: Soft Actor-Critic (SAC) has achieved notable success in continuous control tasks but struggles in sparse reward settings, where infrequent rewards make efficient exploration challenging. While novelty-based exploration methods address this issue by encouraging the agent to explore novel states, they are not trivial to apply to SAC. In particular, managing the interaction between novelty-based exploration and SAC's stochastic policy can lead to inefficient exploration and redundant sample collection. In this paper, we propose KEA (Keeping Exploration Alive) which tackles the inefficiencies in balancing exploration strategies when combining SAC with novelty-based exploration. KEA introduces an additional co-behavior agent that works alongside SAC and a switching mechanism to facilitate proactive coordination between exploration strategies from novelty-based exploration and stochastic policy. This coordination allows the agent to maintain stochasticity in high-novelty regions, enhancing exploration efficiency and reducing repeated sample collection. We first analyze this potential issue in a 2D navigation task and then evaluate KEA on sparse reward control tasks from the DeepMind Control Suite. Compared to state-of-the-art novelty-based exploration baselines, our experiments show that KEA significantly improves learning efficiency and robustness in sparse reward setups.
Authors: Yilong Wang, Jiahao Zhang, Tianxiang Zhao, Suhang Wang
Abstract: Despite their impressive predictive performance, GNNs often exhibit poor confidence calibration, i.e., their predicted confidence scores do not accurately reflect true correctness likelihood. This issue raises concerns about their reliability in high-stakes domains such as fraud detection, and risk assessment, where well-calibrated predictions are essential for decision-making. To ensure trustworthy predictions, several GNN calibration methods are proposed. Though they can improve global calibration, our experiments reveal that they often fail to generalize across different node groups, leading to inaccurate confidence in node groups with different degree levels, classes, and local structures. In certain cases, they even degrade calibration compared to the original uncalibrated GNN. To address this challenge, we propose a novel AdvCali framework that adaptively enhances calibration across different node groups. Our method leverages adversarial training to automatically identify mis-calibrated node groups and applies a differentiable Group Expected Calibration Error (ECE) loss term to refine confidence estimation within these groups. This allows the model to dynamically adjust its calibration strategy without relying on dataset-specific prior knowledge about miscalibrated subgroups. Extensive experiments on real-world datasets demonstrate that our approach not only improves global calibration but also significantly enhances calibration within groups defined by feature similarity, topology, and connectivity, outperforming previous methods and demonstrating its effectiveness in practical scenarios.
Authors: Wei Huang, Hanchen Wang, Dong Wen, Wenjie Zhang, Ying Zhang, Xuemin Lin
Abstract: The Graph Edit Distance (GED) problem, which aims to compute the minimum number of edit operations required to transform one graph into another, is a fundamental challenge in graph analysis with wide-ranging applications. However, due to its NP-hard nature, traditional A* approaches often suffer from scalability issue, making them computationally intractable for large graphs. Many recent deep learning frameworks address GED by formulating it as a regression task, which, while efficient, fails to recover the edit path -- a central interest in GED. Furthermore, recent hybrid approaches that combine deep learning with traditional methods to recover the edit path often yield poor solution quality. These methods also struggle to generate candidate solutions in parallel, resulting in increased running times.In this paper, we present a novel approach, DiffGED, that leverages generative diffusion model to solve GED and recover the corresponding edit path. Specifically, we first generate multiple diverse node matching matrices in parallel through a diffusion-based graph matching model. Next, node mappings are extracted from each generated matching matrices in parallel, and each extracted node mapping can be simply transformed into an edit path. Benefiting from the generative diversity provided by the diffusion model, DiffGED is less likely to fall into local sub-optimal solutions, thereby achieving superior overall solution quality close to the exact solution. Experimental results on real-world datasets demonstrate that DiffGED can generate multiple diverse edit paths with exceptionally high accuracy comparable to exact solutions while maintaining a running time shorter than most of hybrid approaches.
Authors: Varun Mulchandani, Jung-Eun Kim
Abstract: Deep neural networks have been shown to learn and rely on spurious correlations present in the data that they are trained on. Reliance on such correlations can cause these networks to malfunction when deployed in the real world, where these correlations may no longer hold. To overcome the learning of and reliance on such correlations, recent studies propose approaches that yield promising results. These works, however, study settings where the strength of the spurious signal is significantly greater than that of the core, invariant signal, making it easier to detect the presence of spurious features in individual training samples and allow for further processing. In this paper, we identify new settings where the strength of the spurious signal is relatively weaker, making it difficult to detect any spurious information while continuing to have catastrophic consequences. We also discover that spurious correlations are learned primarily due to only a handful of all the samples containing the spurious feature and develop a novel data pruning technique that identifies and prunes small subsets of the training data that contain these samples. Our proposed technique does not require inferred domain knowledge, information regarding the sample-wise presence or nature of spurious information, or human intervention. Finally, we show that such data pruning attains state-of-the-art performance on previously studied settings where spurious information is identifiable.
Authors: Praveen Chopra, Himanshu Kumar, Sandeep Yadav
Abstract: Fault detection in rotating machinery is a complex task, particularly in small and heterogeneous dataset scenarios. Variability in sensor placement, machinery configurations, and structural differences further increase the complexity of the problem. Conventional deep learning approaches often demand large, homogeneous datasets, limiting their applicability in data-scarce industrial environments. While transfer learning and few-shot learning have shown potential, however, they are often constrained by the need for extensive fault datasets. This research introduces a unified framework leveraging a novel progressive neural network (PNN) architecture designed to address these challenges. The PNN sequentially estimates the fixed-size refined features of the higher order with the help of all previously estimated features and appends them to the feature set. This fixed-size feature output at each layer controls the complexity of the PNN and makes it suitable for effective learning from small datasets. The framework's effectiveness is validated on eight datasets, including six open-source datasets, one in-house fault simulator, and one real-world industrial dataset. The PNN achieves state-of-the-art performance in fault detection across varying dataset sizes and machinery types, highlighting superior generalization and classification capabilities.
Authors: Raza Ul Mustafa, Roi Dupart, Gabrielle Smith, Noman Ashraf, Nathalie Japkowicz
Abstract: Islamophobia started evolving into a global phenomenon by attracting followers across the globe, particularly in Western societies. Thus, understanding Islamophobia's global spread and online dissemination is crucial. This paper performs a large-scale analysis of specialized, semi-coded Islamophobic terms such as (muzrat, pislam, mudslime, mohammedan, muzzies) floated on extremist social platforms, i.e., 4Chan, Gab, Telegram, etc. First, we use large language models (LLMs) to show their ability to understand these terms. Second, using Google Perspective API, we also find that Islamophobic text is more toxic compared to other kinds of hate speech. Finally, we use BERT topic modeling approach to extract different topics and Islamophobic discourse on these social platforms. Our findings indicate that LLMs understand these Out-Of-Vocabulary (OOV) slurs; however, measures are still required to control such discourse. Our topic modeling also indicates that Islamophobic text is found across various political, conspiratorial, and far-right movements and is particularly directed against Muslim immigrants. Taken altogether, we performed the first study on Islamophobic semi-coded terms and shed a global light on Islamophobia.
Authors: Christoforos N. Spartalis, Theodoros Semertzidis, Stratis Gavves, Petros Daras
Abstract: We present LoTUS, a novel Machine Unlearning (MU) method that eliminates the influence of training samples from pre-trained models, avoiding retraining from scratch. LoTUS smooths the prediction probabilities of the model -- up to an information theoretic bound -- mitigating its over-confidence that stems from data memorization. We evaluate LoTUS on the Transformer and ResNet18 models, against eight baseline methods, on five public datasets. Beyond established MU benchmarks, we evaluate unlearning on a large-scale dataset (ImageNet1k) which deters retraining, simulating real-world conditions. Moreover, we introduce the novel Retrain-Free Jensen-Shannon Divergence (RF-JSD) metric to enable evaluation under real-world conditions. Experimental results show that LoTUS outperforms state-of-the-art methods in terms of both efficiency and effectiveness. Code: https://github.com/cspartalis/LoTUS.
Authors: Ruijia Zhang, Mingxi Lei, Meng Ding, Zihang Xiang, Jinhui Xu, Di Wang
Abstract: In this paper, we study the problem of (finite sum) minimax optimization in the Differential Privacy (DP) model. Unlike most of the previous studies on the (strongly) convex-concave settings or loss functions satisfying the Polyak-Lojasiewicz condition, here we mainly focus on the nonconvex-strongly-concave one, which encapsulates many models in deep learning such as deep AUC maximization. Specifically, we first analyze a DP version of Stochastic Gradient Descent Ascent (SGDA) and show that it is possible to get a DP estimator whose $l_2$-norm of the gradient for the empirical risk function is upper bounded by $\tilde{O}(\frac{d^{1/4}}{({n\epsilon})^{1/2}})$, where $d$ is the model dimension and $n$ is the sample size. We then propose a new method with less gradient noise variance and improve the upper bound to $\tilde{O}(\frac{d^{1/3}}{(n\epsilon)^{2/3}})$, which matches the best-known result for DP Empirical Risk Minimization with non-convex loss. We also discussed several lower bounds of private minimax optimization. Finally, experiments on AUC maximization, generative adversarial networks, and temporal difference learning with real-world data support our theoretical analysis.
Authors: Wen Zheng Terence Ng, Jianda Chen, Yuan Xu, Tianwei Zhang
Abstract: This work addresses the challenge of personalizing trajectories generated in automated decision-making systems by introducing a resource-efficient approach that enables rapid adaptation to individual users' preferences. Our method leverages a pretrained conditional diffusion model with Preference Latent Embeddings (PLE), trained on a large, reward-free offline dataset. The PLE serves as a compact representation for capturing specific user preferences. By adapting the pretrained model using our proposed preference inversion method, which directly optimizes the learnable PLE, we achieve superior alignment with human preferences compared to existing solutions like Reinforcement Learning from Human Feedback (RLHF) and Low-Rank Adaptation (LoRA). To better reflect practical applications, we create a benchmark experiment using real human preferences on diverse, high-reward trajectories.
Authors: Yunhao Quan, Chuang Gao, Nan Cheng, Zhijie Zhang, Zhisheng Yin, Wenchao Xu, Danyang Wang
Abstract: In Automatic Modulation Classification (AMC), deep learning methods have shown remarkable performance, offering significant advantages over traditional approaches and demonstrating their vast potential. Nevertheless, notable drawbacks, particularly in their high demands for storage, computational resources, and large-scale labeled data, which limit their practical application in real-world scenarios. To tackle this issue, this paper innovatively proposes an automatic modulation classification model based on the Adaptive Lightweight Wavelet Neural Network (ALWNN) and the few-shot framework (MALWNN). The ALWNN model, by integrating the adaptive wavelet neural network and depth separable convolution, reduces the number of model parameters and computational complexity. The MALWNN framework, using ALWNN as an encoder and incorporating prototype network technology, decreases the model's dependence on the quantity of samples. Simulation results indicate that this model performs remarkably well on mainstream datasets. Moreover, in terms of Floating Point Operations Per Second (FLOPS) and Normalized Multiply - Accumulate Complexity (NMACC), ALWNN significantly reduces computational complexity compared to existing methods. This is further validated by real-world system tests on USRP and Raspberry Pi platforms. Experiments with MALWNN show its superior performance in few-shot learning scenarios compared to other algorithms.
Authors: Chang Gao, Kang Zhao, Jianfei Chen, Liping Jing
Abstract: Large language models (LLMs) have demonstrated impressive capabilities, but their enormous size poses significant challenges for deployment in real-world applications. To address this issue, researchers have sought to apply network pruning techniques to LLMs. A critical challenge in pruning is allocation the sparsity for each layer. Recent sparsity allocation methods is often based on heuristics or search that can easily lead to suboptimal performance. In this paper, we conducted an extensive investigation into various LLMs and revealed three significant discoveries: (1) the layerwise pruning sensitivity (LPS) of LLMs is highly non-uniform, (2) the choice of pruning metric affects LPS, and (3) the performance of a sparse model is related to the uniformity of its layerwise redundancy level. Based on these observations, we propose that the layerwise sparsity of LLMs should adhere to three principles: \emph{non-uniformity}, \emph{pruning metric dependency}, and \emph{uniform layerwise redundancy level} in the pruned model. To this end, we proposed Maximum Redundancy Pruning (MRP), an iterative pruning algorithm that prunes in the most redundant layers (\emph{i.e.}, those with the highest non-outlier ratio) at each iteration. The achieved layerwise sparsity aligns with the outlined principles. We conducted extensive experiments on publicly available LLMs, including the LLaMA2 and OPT, across various benchmarks. Experimental results validate the effectiveness of MRP, demonstrating its superiority over previous methods.
Authors: Xudong Mou, Rui Wang, Bo Li, Tianyu Wo, Jie Sun, Hui Wang, Xudong Liu
Abstract: The accumulation of time-series signals and the absence of labels make time-series Anomaly Detection (AD) a self-supervised task of deep learning. Methods based on normality assumptions face the following three limitations: (1) A single assumption could hardly characterize the whole normality or lead to some deviation. (2) Some assumptions may go against the principle of AD. (3) Their basic assumption is that the training data is uncontaminated (free of anomalies), which is unrealistic in practice, leading to a decline in robustness. This paper proposes a novel robust approach, RoCA, which is the first to address all of the above three challenges, as far as we are aware. It fuses the separated assumptions of one-class classification and contrastive learning in a single training process to characterize a more complete so-called normality. Additionally, it monitors the training data and computes a carefully designed anomaly score throughout the training process. This score helps identify latent anomalies, which are then used to define the classification boundary, inspired by the concept of outlier exposure. The performance on AIOps datasets improved by 6% compared to when contamination was not considered (COCA). On two large and high-dimensional multivariate datasets, the performance increased by 5% to 10%. RoCA achieves the highest average performance on both univariate and multivariate datasets. The source code is available at https://github.com/ruiking04/RoCA.
Authors: Siddharth Chandak, Shaan Ul Haque, Nicholas Bambos
Abstract: Two-time-scale Stochastic Approximation (SA) is an iterative algorithm with applications in reinforcement learning and optimization. Prior finite time analysis of such algorithms has focused on fixed point iterations with mappings contractive under Euclidean norm. Motivated by applications in reinforcement learning, we give the first mean square bound on non linear two-time-scale SA where the iterations have arbitrary norm contractive mappings and Markovian noise. We show that the mean square error decays at a rate of $O(1/n^{2/3})$ in the general case, and at a rate of $O(1/n)$ in a special case where the slower timescale is noiseless. Our analysis uses the generalized Moreau envelope to handle the arbitrary norm contractions and solutions of Poisson equation to deal with the Markovian noise. By analyzing the SSP Q-Learning algorithm, we give the first $O(1/n)$ bound for an algorithm for asynchronous control of MDPs under the average reward criterion. We also obtain a rate of $O(1/n)$ for Q-Learning with Polyak-averaging and provide an algorithm for learning Generalized Nash Equilibrium (GNE) for strongly monotone games which converges at a rate of $O(1/n^{2/3})$.
Authors: Kun Li, Xinwei Chen, Tianyou Song, Chengrui Zhou, Zhuoran Liu, Zhenyan Zhang, Jiangjian Guo, Qing Shan
Abstract: In recent years, large language models (LLMs) have shown an impressive ability to perform arithmetic and symbolic reasoning tasks. However, we found that LLMs (e.g., ChatGPT) cannot perform well on reasoning that requires multiple rounds of dialogue, especially when solving situation puzzles. Specifically, LLMs intend to ask very detailed questions focusing on a specific aspect or same/similar questions after several rounds of Q&As. To help LLMs get out of the above dilemma, we propose a novel external reformulation methodology, where the situation puzzle will be reformulated after several rounds of Q&A or when the LLMs raise an incorrect guess. Experiments show superior performance (e.g., win rate, number of question/guess attempts) of our method than directly using LLMs for solving situation puzzles, highlighting the potential of strategic problem reformulation to enhance the reasoning capabilities of LLMs in complex interactive scenarios.
Authors: Wen Bai, Yi Wong, Xiao Qiao, Chin Pang Ho
Abstract: Federated learning (FL) aims to train machine learning (ML) models collaboratively using decentralized data, bypassing the need for centralized data aggregation. Standard FL models often assume that all data come from the same unknown distribution. However, in practical situations, decentralized data frequently exhibit heterogeneity. We propose a novel FL model, Distributionally Robust Federated Learning (DRFL), that applies distributionally robust optimization to overcome the challenges posed by data heterogeneity and distributional ambiguity. We derive a tractable reformulation for DRFL and develop a novel solution method based on the alternating direction method of multipliers (ADMM) algorithm to solve this problem. Our experimental results demonstrate that DRFL outperforms standard FL models under data heterogeneity and ambiguity.
Authors: Tadeusz Dziarmaga, Marcin K\k{a}dzio{\l}ka, Artur Kasymov, Marcin Mazur
Abstract: Deep generative models (DGMs) have caused a paradigm shift in the field of machine learning, yielding noteworthy advancements in domains such as image synthesis, natural language processing, and other related areas. However, a comprehensive evaluation of these models that accounts for the trichotomy between fidelity, diversity, and novelty in generated samples remains a formidable challenge. A recently introduced solution that has emerged as a promising approach in this regard is the Feature Likelihood Divergence (FLD), a method that offers a theoretically motivated practical tool, yet also exhibits some computational challenges. In this paper, we propose PALATE, a novel enhancement to the evaluation of DGMs that addresses limitations of existing metrics. Our approach is based on a peculiar application of the law of total expectation to random variables representing accessible real data. When combined with the MMD baseline metric and DINOv2 feature extractor, PALATE offers a holistic evaluation framework that matches or surpasses state-of-the-art solutions while providing superior computational efficiency and scalability to large-scale datasets. Through a series of experiments, we demonstrate the effectiveness of the PALATE enhancement, contributing a computationally efficient, holistic evaluation approach that advances the field of DGMs assessment, especially in detecting sample memorization and evaluating generalization capabilities.
Authors: Stefan Rass, Martin Dallinger
Abstract: Artificial intelligence models trained from data can only be as good as the underlying data is. Biases in training data propagating through to the output of a machine learning model are a well-documented and well-understood phenomenon, but the machinery to prevent these undesired effects is much less developed. Efforts to ensure data is clean during collection, such as using bias-aware sampling, are most effective when the entity controlling data collection also trains the AI. In cases where the data is already available, how do we find out if the data was already manipulated, i.e., ``poisoned'', so that an undesired behavior would be trained into a machine learning model? This is a challenge fundamentally different to (just) improving approximation accuracy or efficiency, and we provide a method to test training data for flaws, to establish a trustworthy ground-truth for a subsequent training of machine learning models (of any kind). Unlike the well-studied problem of approximating data using fuzzy rules that are generated from the data, our method hinges on a prior definition of rules to happen before seeing the data to be tested. Therefore, the proposed method can also discover hidden error patterns, which may also have substantial influence. Our approach extends the abilities of conventional statistical testing by letting the ``test-condition'' be any Boolean condition to describe a pattern in the data, whose presence we wish to determine. The method puts fuzzy inference into a regression model, to get the best of the two: explainability from fuzzy logic with statistical properties and diagnostics from the regression, and finally also being applicable to ``small data'', hence not requiring large datasets as deep learning methods do. We provide an open source implementation for demonstration and experiments.
Authors: Jiate Li, Meng Pang, Yun Dong, Binghui Wang
Abstract: Graph neural networks (GNNs) are becoming the de facto method to learn on the graph data and have achieved the state-of-the-art on node and graph classification tasks. However, recent works show GNNs are vulnerable to training-time poisoning attacks -- marginally perturbing edges, nodes, or/and node features of training graph(s) can largely degrade GNNs' testing performance. Most previous defenses against graph poisoning attacks are empirical and are soon broken by adaptive / stronger ones. A few provable defenses provide robustness guarantees, but have large gaps when applied in practice: 1) restrict the attacker on only one type of perturbation; 2) design for a particular GNN architecture or task; and 3) robustness guarantees are not 100\% accurate. In this work, we bridge all these gaps by developing PGNNCert, the first certified defense of GNNs against poisoning attacks under arbitrary (edge, node, and node feature) perturbations with deterministic robustness guarantees. Extensive evaluations on multiple node and graph classification datasets and GNNs demonstrate the effectiveness of PGNNCert to provably defend against arbitrary poisoning perturbations. PGNNCert is also shown to significantly outperform the state-of-the-art certified defenses against edge perturbation or node perturbation during GNN training.
Authors: Fei Zhu, Yujing Liu, Wenzhuo Liu, Zhaoxiang Zhang
Abstract: Continual learning, which aims to learn multiple tasks sequentially, has gained extensive attention. However, most existing work focuses on empirical studies, and the theoretical aspect remains under-explored. Recently, a few investigations have considered the theory of continual learning only for linear regressions, establishes the results based on the strict independent and identically distributed (i.i.d.) assumption and the persistent excitation on the feature data that may be difficult to verify or guarantee in practice. To overcome this fundamental limitation, in this paper, we provide a general and comprehensive theoretical analysis for continual learning of regression models. By utilizing the stochastic Lyapunov function and martingale estimation techniques, we establish the almost sure convergence results of continual learning under a general data condition for the first time. Additionally, without any excitation condition imposed on the data, the convergence rates for the forgetting and regret metrics are provided.
Authors: Moein Sorkhei, Christos Matsoukas, Johan Fredin Haslum, Kevin Smith
Abstract: How well can one expect transfer learning to work in a new setting where the domain is shifted, the task is different, and the architecture changes? Many transfer learning metrics have been proposed to answer this question. But how accurate are their predictions in a realistic new setting? We conducted an extensive evaluation involving over 42,000 experiments comparing 23 transferability metrics across 16 different datasets to assess their ability to predict transfer performance. Our findings reveal that none of the existing metrics perform well across the board. However, we find that a simple k-nearest neighbor evaluation -- as is commonly used to evaluate feature quality for self-supervision -- not only surpasses existing metrics, but also offers better computational efficiency and ease of implementation.
Authors: Xiaolong Yin, Xingyu Lu, Jiahang Shen, Jingzhe Ni, Hailong Li, Ruofeng Tong, Min Tang, Peng Du
Abstract: A CAD command sequence is a typical parametric design paradigm in 3D CAD systems where a model is constructed by overlaying 2D sketches with operations such as extrusion, revolution, and Boolean operations. Although there is growing academic interest in the automatic generation of command sequences, existing methods and datasets only support operations such as 2D sketching, extrusion,and Boolean operations. This limitation makes it challenging to represent more complex geometries. In this paper, we present a reinforcement learning (RL) training environment (gym) built on a CAD geometric engine. Given an input boundary representation (B-Rep) geometry, the policy network in the RL algorithm generates an action. This action, along with previously generated actions, is processed within the gym to produce the corresponding CAD geometry, which is then fed back into the policy network. The rewards, determined by the difference between the generated and target geometries within the gym, are used to update the RL network. Our method supports operations beyond sketches, Boolean, and extrusion, including revolution operations. With this training gym, we achieve state-of-the-art (SOTA) quality in generating command sequences from B-Rep geometries. In addition, our method can significantly improve the efficiency of command sequence generation by a factor of 39X compared with the previous training gym.
Authors: Eoin Quinn, Ghassene Jebali, Maxime Seince, Oliver Bent
Abstract: We explore a framework for protein sequence representation learning that decomposes the task between manifold learning and distributional modelling. Specifically we present a Latent Space Diffusion architecture which combines a protein sequence autoencoder with a denoising diffusion model operating on its latent space. We obtain a one-parameter family of learned representations from the diffusion model, along with the autoencoder's latent representation. We propose and evaluate two autoencoder architectures: a homogeneous model forcing amino acids of the same type to be identically distributed in the latent space, and an inhomogeneous model employing a noise-based variant of masking. As a baseline we take a latent space learned by masked language modelling, and evaluate discriminative capability on a range of protein property prediction tasks. Our finding is twofold: the diffusion models trained on both our proposed variants display higher discriminative power than the one trained on the masked language model baseline, none of the diffusion representations achieve the performance of the masked language model embeddings themselves.
Authors: Abdoul Majid O. Thiombiano, Brahim Hnich, Ali Ben Mrad, Mohamed Wiem Mkaouer
Abstract: The current era of Natural Language Processing (NLP) is dominated by Transformer models. However, novel architectures relying on recurrent mechanisms, such as xLSTM and Mamba, have been proposed as alternatives to attention-based models. Although computation is done differently than with the attention mechanism mechanism, these recurrent models yield good results and sometimes even outperform state-of-the-art attention-based models. In this work, we propose Distil-xLSTM, an xLSTM-based Small Language Model (SLM) trained by distilling knowledge from a Large Language Model (LLM) that shows promising results while being compute and scale efficient. Our Distil-xLSTM focuses on approximating a transformer-based model attention parametrization using its recurrent sequence mixing components and shows good results with minimal training.
Authors: Hadi Mohammadi, Ehsan Nazerfard, Mostafa Haghir Chehreghani
Abstract: Imbalanced data represent a distribution with more frequencies of one class (majority) than the other (minority). This phenomenon occurs across various domains, such as security, medical care and human activity. In imbalanced learning, classification algorithms are typically inclined to classify the majority class accurately, resulting in artificially high accuracy rates. As a result, many minority samples are mistakenly labelled as majority-class instances, resulting in a bias that benefits the majority class. This study presents a framework based on boundary anchor samples to tackle the imbalance learning challenge. First, we select and use anchor samples to train a multilayer perceptron (MLP) classifier, which acts as a prior knowledge model and aids the adversarial and contrastive learning procedures. Then, we designed a novel deep generative model called Anchor Stabilized Conditional Generative Adversarial Network or Anch-SCGAN in short. Anch-SCGAN is supported with two generators for the minority and majority classes and a discriminator incorporating additional class-specific information from the pre-trained feature extractor MLP. In addition, we facilitate the generator's training procedure in two ways. First, we define a new generator loss function based on reprocessed anchor samples and contrastive learning. Second, we apply a scoring strategy to stabilize the adversarial training part in generators. We train Anch-SCGAN and further finetune it with anchor samples to improve the precision of the generated samples. Our experiments on 16 real-world imbalanced datasets illustrate that Anch-SCGAN outperforms the renowned methods in imbalanced learning.
Authors: Tianyu Chen, Xingcheng Fu, Yisen Gao, Haodong Qian, Yuecen Wei, Kun Yan, Haoyi Zhou, Jianxin Li
Abstract: Modern vision-language models (VLMs) develop patch embedding and convolution backbone within vector space, especially Euclidean ones, at the very founding. When expanding VLMs to a galaxy scale for understanding astronomical phenomena, the integration of spherical space for planetary orbits and hyperbolic spaces for black holes raises two formidable challenges. a) The current pre-training model is confined to Euclidean space rather than a comprehensive geometric embedding. b) The predominant architecture lacks suitable backbones for anisotropic physical geometries. In this paper, we introduced Galaxy-Walker, a geometry-aware VLM, for the universe-level vision understanding tasks. We proposed the geometry prompt that generates geometry tokens by random walks across diverse spaces on a multi-scale physical graph, along with a geometry adapter that compresses and reshapes the space anisotropy in a mixture-of-experts manner. Extensive experiments demonstrate the effectiveness of our approach, with Galaxy-Walker achieving state-of-the-art performance in both galaxy property estimation ($R^2$ scores up to $0.91$) and morphology classification tasks (up to $+0.17$ F1 improvement in challenging features), significantly outperforming both domain-specific models and general-purpose VLMs.
Authors: Zhiwei Shi, Chengxi Zhu, Fan Yang, Jun Yan, Zheyun Qin, Songquan Shi, Zhumin Chen
Abstract: This paper presents a data driven universal ball trajectory prediction method integrated with physics equations. Existing methods are designed for specific ball types and struggle to generalize. This challenge arises from three key factors. First, learning-based models require large datasets but suffer from accuracy drops in unseen scenarios. Second, physics-based models rely on complex formulas and detailed inputs, yet accurately obtaining ball states, such as spin, is often impractical. Third, integrating physical principles with neural networks to achieve high accuracy, fast inference, and strong generalization remains difficult. To address these issues, we propose an innovative approach that incorporates physics-based equations and neural networks. We first derive three generalized physical formulas. Then, using a neural network and observed trajectory points, we infer certain parameters while fitting the remaining ones. These formulas enable precise trajectory prediction with minimal training data: only a few dozen samples. Extensive experiments demonstrate our method superiority in generalization, real-time performance, and accuracy.
Authors: Chengxiang Huang, Yake Wei, Zequn Yang, Di Hu
Abstract: Sensory training during the early ages is vital for human development. Inspired by this cognitive phenomenon, we observe that the early training stage is also important for the multimodal learning process, where dataset information is rapidly acquired. We refer to this stage as the prime learning window. However, based on our observation, this prime learning window in multimodal learning is often dominated by information-sufficient modalities, which in turn suppresses the information acquisition of information-insufficient modalities. To address this issue, we propose Information Acquisition Regulation (InfoReg), a method designed to balance information acquisition among modalities. Specifically, InfoReg slows down the information acquisition process of information-sufficient modalities during the prime learning window, which could promote information acquisition of information-insufficient modalities. This regulation enables a more balanced learning process and improves the overall performance of the multimodal network. Experiments show that InfoReg outperforms related multimodal imbalanced methods across various datasets, achieving superior model performance. The code is available at https://github.com/GeWu-Lab/InfoReg_CVPR2025.
Authors: Mohsen Amiri, Sindri Magn\'usson
Abstract: Reinforcement learning in non-stationary environments is challenging due to abrupt and unpredictable changes in dynamics, often causing traditional algorithms to fail to converge. However, in many real-world cases, non-stationarity has some structure that can be exploited to develop algorithms and facilitate theoretical analysis. We introduce one such structure, Switching Non-Stationary Markov Decision Processes (SNS-MDP), where environments switch over time based on an underlying Markov chain. Under a fixed policy, the value function of an SNS-MDP admits a closed-form solution determined by the Markov chain's statistical properties, and despite the inherent non-stationarity, Temporal Difference (TD) learning methods still converge to the correct value function. Furthermore, policy improvement can be performed, and it is shown that policy iteration converges to the optimal policy. Moreover, since Q-learning converges to the optimal Q-function, it likewise yields the corresponding optimal policy. To illustrate the practical advantages of SNS-MDPs, we present an example in communication networks where channel noise follows a Markovian pattern, demonstrating how this framework can effectively guide decision-making in complex, time-varying contexts.
Authors: Yongshuai Liu, Xin Liu
Abstract: Recent developments in deep reinforcement learning have been very successful in learning complex, previously intractable problems. Sample efficiency and local optimality, however, remain significant challenges. To address these challenges, novelty-driven exploration strategies have emerged and shown promising potential. Unfortunately, no single algorithm outperforms all others in all tasks and most of them struggle with tasks with high-dimensional and complex observations. In this work, we propose Adventurer, a novelty-driven exploration algorithm that is based on Bidirectional Generative Adversarial Networks (BiGAN), where BiGAN is trained to estimate state novelty. Intuitively, a generator that has been trained on the distribution of visited states should only be able to generate a state coming from the distribution of visited states. As a result, novel states using the generator to reconstruct input states from certain latent representations would lead to larger reconstruction errors. We show that BiGAN performs well in estimating state novelty for complex observations. This novelty estimation method can be combined with intrinsic-reward-based exploration. Our empirical results show that Adventurer produces competitive results on a range of popular benchmark tasks, including continuous robotic manipulation tasks (e.g. Mujoco robotics) and high-dimensional image-based tasks (e.g. Atari games).
Authors: Sebasti\'an A. Cajas Ord\'o\~nez, Jaydeep Samanta, Andr\'es L. Su\'arez-Cetrulo, Ricardo Sim\'on Carbajo
Abstract: The Internet of Things is an example domain where data is perpetually generated in ever-increasing quantities, reflecting the proliferation of connected devices and the formation of continuous data streams over time. Consequently, the demand for ad-hoc, cost-effective machine learning solutions must adapt to this evolving data influx. This study tackles the task of offloading in small gateways, exacerbated by their dynamic availability over time. An approach leveraging CPU utilization metrics using online and continual machine learning techniques is proposed to predict gateway availability. These methods are compared to popular machine learning algorithms and a recent time-series foundation model, Lag-Llama, for fine-tuned and zero-shot setups. Their performance is benchmarked on a dataset of CPU utilization measurements over time from an IoT gateway and focuses on model metrics such as prediction errors, training and inference times, and memory consumption. Our primary objective is to study new efficient ways to predict CPU performance in IoT environments. Across various scenarios, our findings highlight that ensemble and online methods offer promising results for this task in terms of accuracy while maintaining a low resource footprint.
Authors: Aditya Sai Ellendula, Arun K Pujari, Vikas Kumar
Abstract: This paper presents an efficient preference elicitation framework for uncertain matroid optimization, where precise weight information is unavailable, but insights into possible weight values are accessible. The core innovation of our approach lies in its ability to systematically elicit user preferences, aligning the optimization process more closely with decision-makers' objectives. By incrementally querying preferences between pairs of elements, we iteratively refine the parametric uncertainty regions, leveraging the structural properties of matroids. Our method aims to achieve the exact optimum by reducing regret with a few elicitation rounds. Additionally, our approach avoids the computation of Minimax Regret and the use of Linear programming solvers at every iteration, unlike previous methods. Experimental results on four standard matroids demonstrate that our method reaches optimality more quickly and with fewer preference queries than existing techniques.
Authors: Feilong Cao, Shao-Bo Lin
Abstract: The great success of deep learning has stimulated avid research activities in verifying the power of depth in theory, a common consensus of which is that deep net are versatile in approximating and learning numerous functions. Such a versatility certainly enhances the understanding of the power of depth, but makes it difficult to judge which data features are crucial in a specific learning task. This paper proposes a constructive approach to equip deep nets for the feature qualification purpose. Using the product-gate nature and localized approximation property of deep nets with sigmoid activation (deep sigmoid nets), we succeed in constructing a linear deep net operator that possesses optimal approximation performance in approximating smooth and radial functions. Furthermore, we provide theoretical evidences that the constructed deep net operator is capable of qualifying multiple features such as the smoothness and radialness of the target functions.
Authors: Changho Shin, Xinya Yan, Suenggwan Jo, Sungjun Cho, Shourjo Aditya Chaudhuri, Frederic Sala
Abstract: Language models often struggle with temporal misalignment, performance degradation caused by shifts in the temporal distribution of data. Continuously updating models to avoid degradation is expensive. Can models be adapted without updating model weights? We present TARDIS, an unsupervised representation editing method that addresses this challenge. TARDIS extracts steering vectors from unlabeled data and adjusts the model's representations to better align with the target time period's distribution. Our experiments reveal that TARDIS enhances downstream task performance without the need for fine-tuning, can mitigate temporal misalignment even when exact target time period data is unavailable, and remains efficient even when the temporal information of the target data points is unknown at inference time.
Authors: Chris Pedersen, Laure Zanna, Joan Bruna
Abstract: Autoregressive surrogate models (or \textit{emulators}) of spatiotemporal systems provide an avenue for fast, approximate predictions, with broad applications across science and engineering. At inference time, however, these models are generally unable to provide predictions over long time rollouts due to accumulation of errors leading to diverging trajectories. In essence, emulators operate out of distribution, and controlling the online distribution quickly becomes intractable in large-scale settings. To address this fundamental issue, and focusing on time-stationary systems admitting an invariant measure, we leverage diffusion models to obtain an implicit estimator of the score of this invariant measure. We show that this model of the score function can be used to stabilize autoregressive emulator rollouts by applying on-the-fly denoising during inference, a process we call \textit{thermalization}. Thermalizing an emulator rollout is shown to extend the time horizon of stable predictions by an order of magnitude in complex systems exhibiting turbulent and chaotic behavior, opening up a novel application of diffusion models in the context of neural emulation.
Authors: Florian Rupp, Manuel Eberhardinger, Kai Eckert
Abstract: The balancing process for game levels in competitive two-player contexts involves a lot of manual work and testing, particularly for non-symmetrical game levels. In this work, we frame game balancing as a procedural content generation task and propose an architecture for automatically balancing of tile-based levels within the PCGRL framework (procedural content generation via reinforcement learning). Our architecture is divided into three parts: (1) a level generator, (2) a balancing agent, and (3) a reward modeling simulation. Through repeated simulations, the balancing agent receives rewards for adjusting the level towards a given balancing objective, such as equal win rates for all players. To this end, we propose new swap-based representations to improve the robustness of playability, thereby enabling agents to balance game levels more effectively and quickly compared to traditional PCGRL. By analyzing the agent's swapping behavior, we can infer which tile types have the most impact on the balance. We validate our approach in the Neural MMO (NMMO) environment in a competitive two-player scenario. In this extended conference paper, we present improved results, explore the applicability of the method to various forms of balancing beyond equal balancing, compare the performance to another search-based approach, and discuss the application of existing fairness metrics to game balancing.
Authors: Nooshin Bahador
Abstract: Mechanistic interpretability improves the safety, reliability, and robustness of large AI models. This study examined individual attention heads in vision transformers (ViTs) fine tuned on distorted 2D spectrogram images containing non relevant content (axis labels, titles, color bars). By introducing extraneous features, the study analyzed how transformer components processed unrelated information, using mechanistic interpretability to debug issues and reveal insights into transformer architectures. Attention maps assessed head contributions across layers. Heads in early layers (1 to 3) showed minimal task impact with ablation increased MSE loss slightly ({\mu}=0.11%, {\sigma}=0.09%), indicating focus on less critical low level features. In contrast, deeper heads (e.g., layer 6) caused a threefold higher loss increase ({\mu}=0.34%, {\sigma}=0.02%), demonstrating greater task importance. Intermediate layers (6 to 11) exhibited monosemantic behavior, attending exclusively to chirp regions. Some early heads (1 to 4) were monosemantic but non task relevant (e.g. text detectors, edge or corner detectors). Attention maps distinguished monosemantic heads (precise chirp localization) from polysemantic heads (multiple irrelevant regions). These findings revealed functional specialization in ViTs, showing how heads processed relevant vs. extraneous information. By decomposing transformers into interpretable components, this work enhanced model understanding, identified vulnerabilities, and advanced safer, more transparent AI.
Authors: Daniel Mayfrank, Mehmet Velioglu, Alexander Mitsos, Manuel Dahmen
Abstract: Reinforcement learning (RL) can be used to tune data-driven (economic) nonlinear model predictive controllers ((e)NMPCs) for optimal performance in a specific control task by optimizing the dynamic model or parameters in the policy's objective function or constraints, such as state bounds. However, the sample efficiency of RL is crucial, and to improve it, we combine a model-based RL algorithm with our published method that turns Koopman (e)NMPCs into automatically differentiable policies. We apply our approach to an eNMPC case study of a continuous stirred-tank reactor (CSTR) model from the literature. The approach outperforms benchmark methods, i.e., data-driven eNMPCs using models based on system identification without further RL tuning of the resulting policy, and neural network controllers trained with model-based RL, by achieving superior control performance and higher sample efficiency. Furthermore, utilizing partial prior knowledge about the system dynamics via physics-informed learning further increases sample efficiency.
Authors: Tan-Khiem Huynh, Malcolm Egan, Giovanni Neglia, Jean-Marie Gorce
Abstract: Federated learning (FL) is now recognized as a key framework for communication-efficient collaborative learning. Most theoretical and empirical studies, however, rely on the assumption that clients have access to pre-collected data sets, with limited investigation into scenarios where clients continuously collect data. In many real-world applications, particularly when data is generated by physical or biological processes, client data streams are often modeled by non-stationary Markov processes. Unlike standard i.i.d. sampling, the performance of FL with Markovian data streams remains poorly understood due to the statistical dependencies between client samples over time. In this paper, we investigate whether FL can still support collaborative learning with Markovian data streams. Specifically, we analyze the performance of Minibatch SGD, Local SGD, and a variant of Local SGD with momentum. We answer affirmatively under standard assumptions and smooth non-convex client objectives: the sample complexity is proportional to the inverse of the number of clients with a communication complexity comparable to the i.i.d. scenario. However, the sample complexity for Markovian data streams remains higher than for i.i.d. sampling.
Authors: Daphne Lenders, Andrea Pugnana, Roberto Pellungrini, Toon Calders, Dino Pedreschi, Fosca Giannotti
Abstract: Abstaining classifiers have the option to refrain from providing a prediction for instances that are difficult to classify. The abstention mechanism is designed to trade off the classifier's performance on the accepted data while ensuring a minimum number of predictions. In this setting, often fairness concerns arise when the abstention mechanism solely reduces errors for the majority groups of the data, resulting in increased performance differences across demographic groups. While there exist a bunch of methods that aim to reduce discrimination when abstaining, there is no mechanism that can do so in an explainable way. In this paper, we fill this gap by introducing Interpretable and Fair Abstaining Classifier IFAC, an algorithm that can reject predictions both based on their uncertainty and their unfairness. By rejecting possibly unfair predictions, our method reduces error and positive decision rate differences across demographic groups of the non-rejected data. Since the unfairness-based rejections are based on an interpretable-by-design method, i.e., rule-based fairness checks and situation testing, we create a transparent process that can empower human decision-makers to review the unfair predictions and make more just decisions for them. This explainable aspect is especially important in light of recent AI regulations, mandating that any high-risk decision task should be overseen by human experts to reduce discrimination risks.
Authors: Xuan Li, Yuting Peng, Xiaoxuan Sun, Yifei Duan, Zhou Fang, Tengda Tang
Abstract: With the rapid development of e-commerce, e-commerce platforms are facing an increasing number of fraud threats. Effectively identifying and preventing these fraudulent activities has become a critical research problem. Traditional fraud detection methods typically rely on supervised learning, which requires large amounts of labeled data. However, such data is often difficult to obtain, and the continuous evolution of fraudulent activities further reduces the adaptability and effectiveness of traditional methods. To address this issue, this study proposes an unsupervised e-commerce fraud detection algorithm based on SimCLR. The algorithm leverages the contrastive learning framework to effectively detect fraud by learning the underlying representations of transaction data in an unlabeled setting. Experimental results on the eBay platform dataset show that the proposed algorithm outperforms traditional unsupervised methods such as K-means, Isolation Forest, and Autoencoders in terms of accuracy, precision, recall, and F1 score, demonstrating strong fraud detection capabilities. The results confirm that the SimCLR-based unsupervised fraud detection method has broad application prospects in e-commerce platform security, improving both detection accuracy and robustness. In the future, with the increasing scale and diversity of datasets, the model's performance will continue to improve, and it could be integrated with real-time monitoring systems to provide more efficient security for e-commerce platforms.
Authors: Daniel Lepe-Soltero, Thierry Arti\`eres, Ana\"is Baudot, Paul Villoutreix
Abstract: A key challenge today lies in the ability to efficiently handle multi-omics data since such multimodal data may provide a more comprehensive overview of the underlying processes in a system. Yet it comes with challenges: multi-omics data are most often unpaired and only partially labeled, moreover only small amounts of data are available in some situation such as rare diseases. We propose MODIS which stands for Multi-Omics Data Integration for Small and unpaired datasets, a semi supervised approach to account for these particular settings. MODIS learns a probabilistic coupling of heterogeneous data modalities and learns a shared latent space where modalities are aligned. We rely on artificial data to build controlled experiments to explore how much supervision is needed for an accurate alignment of modalities, and how our approach enables dealing with new conditions for which few data are available. The code is available athttps://github.com/VILLOUTREIXLab/MODIS.
Authors: Yangjun Ruan, Neil Band, Chris J. Maddison, Tatsunori Hashimoto
Abstract: Compute scaling for language model (LM) pretraining has outpaced the growth of human-written texts, leading to concerns that data will become the bottleneck to LM scaling. To continue scaling pretraining in this data-constrained regime, we propose that explicitly modeling and inferring the latent thoughts that underlie the text generation process can significantly improve pretraining data efficiency. Intuitively, our approach views web text as the compressed final outcome of a verbose human thought process and that the latent thoughts contain important contextual knowledge and reasoning steps that are critical to data-efficient learning. We empirically demonstrate the effectiveness of our approach through data-constrained continued pretraining for math. We first show that synthetic data approaches to inferring latent thoughts significantly improve data efficiency, outperforming training on the same amount of raw data (5.7\% $\rightarrow$ 25.4\% on MATH). Furthermore, we demonstrate latent thought inference without a strong teacher, where an LM bootstraps its own performance by using an EM algorithm to iteratively improve the capability of the trained LM and the quality of thought-augmented pretraining data. We show that a 1B LM can bootstrap its performance across at least three iterations and significantly outperform baselines trained on raw data, with increasing gains from additional inference compute when performing the E-step. The gains from inference scaling and EM iterations suggest new opportunities for scaling data-constrained pretraining.
Authors: Yuhang Wang, Hanwei Guo, Sizhe Wang, Long Qian, Xuguang Lan
Abstract: Model Predictive Control (MPC) has been demonstrated to be effective in continuous control tasks. When a world model and a value function are available, planning a sequence of actions ahead of time leads to a better policy. Existing methods typically obtain the value function and the corresponding policy in a model-free manner. However, we find that such an approach struggles with complex tasks, resulting in poor policy learning and inaccurate value estimation. To address this problem, we leverage the strengths of MPC itself. In this work, we introduce Bootstrapped Model Predictive Control (BMPC), a novel algorithm that performs policy learning in a bootstrapped manner. BMPC learns a network policy by imitating an MPC expert, and in turn, uses this policy to guide the MPC process. Combined with model-based TD-learning, our policy learning yields better value estimation and further boosts the efficiency of MPC. We also introduce a lazy reanalyze mechanism, which enables computationally efficient imitation learning. Our method achieves superior performance over prior works on diverse continuous control tasks. In particular, on challenging high-dimensional locomotion tasks, BMPC significantly improves data efficiency while also enhancing asymptotic performance and training stability, with comparable training time and smaller network sizes. Code is available at https://github.com/wertyuilife2/bmpc.
Authors: Runze Cheng, Yao Sun, Lan Zhang, Lei Feng, Lei Zhang, Muhammad Ali Imran
Abstract: With the significant advances in generative AI (GAI) and the proliferation of mobile devices, providing high-quality AI-generated content (AIGC) services via wireless networks is becoming the future direction. However, the primary challenges of AIGC service delivery in wireless networks lie in unstable channels, limited bandwidth resources, and unevenly distributed computational resources. In this paper, we employ semantic communication (SemCom) in diffusion-based GAI models to propose a Resource-aware wOrkload-adjUstable TransceivEr (ROUTE) for AIGC delivery in dynamic wireless networks. Specifically, to relieve the communication resource bottleneck, SemCom is utilized to prioritize semantic information of the generated content. Then, to improve computational resource utilization in both edge and local and reduce AIGC semantic distortion in transmission, modified diffusion-based models are applied to adjust the computing workload and semantic density in cooperative content generation. Simulations verify the superiority of our proposed ROUTE in terms of latency and content quality compared to conventional AIGC approaches.
Authors: Weihao Zeng, Yuzhen Huang, Qian Liu, Wei Liu, Keqing He, Zejun Ma, Junxian He
Abstract: DeepSeek-R1 has shown that long chain-of-thought (CoT) reasoning can naturally emerge through a simple reinforcement learning (RL) framework with rule-based rewards, where the training may directly start from the base models-a paradigm referred to as zero RL training. Most recent efforts to reproduce zero RL training have primarily focused on the Qwen2.5 model series, which may not be representative as we find the base models already exhibit strong instruction-following and self-reflection abilities. In this work, we investigate zero RL training across 10 diverse base models, spanning different families and sizes including LLama3-8B, Mistral-7B/24B, DeepSeek-Math-7B, Qwen2.5-math-7B, and all Qwen2.5 models from 0.5B to 32B. Leveraging several key design strategies-such as adjusting format reward and controlling query difficulty-we achieve substantial improvements in both reasoning accuracy and response length across most settings. However, by carefully monitoring the training dynamics, we observe that different base models exhibit distinct patterns during training. For instance, the increased response length does not always correlate with the emergence of certain cognitive behaviors such as verification (i.e., the "aha moment"). Notably, we observe the "aha moment" for the first time in small models not from the Qwen family. We share the key designs that enable successful zero RL training, along with our findings and practices. To facilitate further research, we open-source the code, models, and analysis tools.
Authors: Akhiad Bercovich, Mohammad Dabbah, Omri Puny, Ido Galil, Amnon Geifman, Yonatan Geifman, Izhak Golan, Ehud Karpas, Itay Levy, Zach Moshe, Najeeb Nabwani, Tomer Ronen, Itamar Schen, Elad Segal, Ido Shahaf, Oren Tropp, Ran Zilberstein, Ran El-Yaniv
Abstract: We introduce FFN Fusion, an architectural optimization technique that reduces sequential computation in large language models by identifying and exploiting natural opportunities for parallelization. Our key insight is that sequences of Feed-Forward Network (FFN) layers, particularly those remaining after the removal of specific attention layers, can often be parallelized with minimal accuracy impact. We develop a principled methodology for identifying and fusing such sequences, transforming them into parallel operations that significantly reduce inference latency while preserving model behavior. Applying these techniques to Llama-3.1-405B-Instruct, we create Llama-Nemotron-Ultra-253B-Base (Ultra-253B-Base), an efficient and soon-to-be publicly available model that achieves a 1.71X speedup in inference latency and 35X lower per-token cost while maintaining strong performance across benchmarks. Through extensive experiments on models from 49B to 253B parameters, we demonstrate that FFN Fusion becomes increasingly effective at larger scales and can complement existing optimization techniques like quantization and pruning. Most intriguingly, we find that even full transformer blocks containing both attention and FFN layers can sometimes be parallelized, suggesting new directions for neural architecture design.
Authors: Brian R. Bartoldson, Siddarth Venkatraman, James Diffenderfer, Moksh Jain, Tal Ben-Nun, Seanie Lee, Minsu Kim, Johan Obando-Ceron, Yoshua Bengio, Bhavya Kailkhura
Abstract: Reinforcement learning (RL) is a critical component of large language model (LLM) post-training. However, existing on-policy algorithms used for post-training are inherently incompatible with the use of experience replay buffers, which can be populated scalably by distributed off-policy actors to enhance exploration as compute increases. We propose efficiently obtaining this benefit of replay buffers via Trajectory Balance with Asynchrony (TBA), a massively scalable LLM RL system. In contrast to existing approaches, TBA uses a larger fraction of compute on search, constantly generating off-policy data for a central replay buffer. A training node simultaneously samples data from this buffer based on reward or recency to update the policy using Trajectory Balance (TB), a diversity-seeking RL objective introduced for GFlowNets. TBA offers three key advantages: (1) decoupled training and search, speeding up training wall-clock time by 4x or more; (2) improved diversity through large-scale off-policy sampling; and (3) scalable search for sparse reward settings. On mathematical reasoning, preference-tuning, and automated red-teaming (diverse and representative post-training tasks), TBA produces speed and performance improvements over strong baselines.
Authors: Jassem Abbasi, Ameya D. Jagtap, Ben Moseley, Aksel Hiorth, P{\aa}l {\O}steb{\o} Andersen
Abstract: Solving partial differential equations (PDEs) with discontinuous solutions , such as shock waves in multiphase viscous flow in porous media , is critical for a wide range of scientific and engineering applications, as they represent sudden changes in physical quantities. Physics-Informed Neural Networks (PINNs), an approach proposed for solving PDEs, encounter significant challenges when applied to such systems. Accurately solving PDEs with discontinuities using PINNs requires specialized techniques to ensure effective solution accuracy and numerical stability. A benchmarking study was conducted on two multiphase flow problems in porous media: the classic Buckley-Leverett (BL) problem and a fully coupled system of equations involving shock waves but with varying levels of solution complexity. The findings show that PM and LM approaches can provide accurate solutions for the BL problem by effectively addressing the infinite gradients associated with shock occurrences. In contrast, AM methods failed to effectively resolve the shock waves. When applied to fully coupled PDEs (with more complex loss landscape), the generalization error in the solutions quickly increased, highlighting the need for ongoing innovation. This study provides a comprehensive review of existing techniques for managing PDE discontinuities using PINNs, offering information on their strengths and limitations. The results underscore the necessity for further research to improve PINNs ability to handle complex discontinuities, particularly in more challenging problems with complex loss landscapes. This includes problems involving higher dimensions or multiphysics systems, where current methods often struggle to maintain accuracy and efficiency.
Authors: Xu Wu, Lesego E. Moloko, Pavel M. Bokov, Gregory K. Delipei, Joshua Kaizer, Kostadin N. Ivanov
Abstract: Machine learning (ML) has been leveraged to tackle a diverse range of tasks in almost all branches of nuclear engineering. Many of the successes in ML applications can be attributed to the recent performance breakthroughs in deep learning, the growing availability of computational power, data, and easy-to-use ML libraries. However, these empirical successes have often outpaced our formal understanding of the ML algorithms. An important but under-rated area is uncertainty quantification (UQ) of ML. ML-based models are subject to approximation uncertainty when they are used to make predictions, due to sources including but not limited to, data noise, data coverage, extrapolation, imperfect model architecture and the stochastic training process. The goal of this paper is to clearly explain and illustrate the importance of UQ of ML. We will elucidate the differences in the basic concepts of UQ of physics-based models and data-driven ML models. Various sources of uncertainties in physical modeling and data-driven modeling will be discussed, demonstrated, and compared. We will also present and demonstrate a few techniques to quantify the ML prediction uncertainties. Finally, we will discuss the need for building a verification, validation and UQ framework to establish ML credibility.
Authors: Haoran Li, Yingxue Zhao, Haosu Zhou, Tobias Pfaff, Nan Li
Abstract: During the design cycle of safety critical vehicle components such as B-pillars, crashworthiness performance is a key metric for passenger protection assessment in vehicle accidents. Traditional finite element simulations for crashworthiness analysis involve complex modelling, leading to an increased computational demand. Although a few machine learning-based surrogate models have been developed for rapid predictions for crashworthiness analysis, they exhibit limitations in detailed representation of complex 3D components. Graph Neural Networks (GNNs) have emerged as a promising solution for processing data with complex structures. However, existing GNN models often lack sufficient accuracy and computational efficiency to meet industrial demands. This paper proposes Recurrent Graph U-Net (ReGUNet), a new graph-based surrogate model for crashworthiness analysis of vehicle panel components. ReGUNet adoptes a U-Net architecture with multiple graph downsampling and upsampling layers, which improves the model's computational efficiency and accuracy; the introduction of recurrence enhances the accuracy and stability of temporal predictions over multiple time steps. ReGUNet is evaluated through a case study of side crash testing of a B-pillar component with variation in geometric design. The trained model demonstrates great accuracy in predicting the dynamic behaviour of previously unseen component designs within a relative error of 0.74% for the maximum B-pillar intrusion. Compared to the baseline models, ReGUNet can reduce the averaged mean prediction error of the component's deformation by more than 51% with significant improvement in computational efficiency. Provided enhanced accuracy and efficiency, ReGUNet shows greater potential in accurate predictions of large and complex graphs compared to existing models.
Authors: Dillon Bowen, Ann-Kathrin Dombrowski, Adam Gleave, Chris Cundy
Abstract: The rapid advancement of AI systems has raised widespread concerns about potential harms of frontier AI systems and the need for responsible evaluation and oversight. In this position paper, we argue that frontier AI companies should report both pre- and post-mitigation safety evaluations to enable informed policy decisions. Evaluating models at both stages provides policymakers with essential evidence to regulate deployment, access, and safety standards. We show that relying on either in isolation can create a misleading picture of model safety. Our analysis of AI safety disclosures from leading frontier labs identifies three critical gaps: (1) companies rarely evaluate both pre- and post-mitigation versions, (2) evaluation methods lack standardization, and (3) reported results are often too vague to inform policy. To address these issues, we recommend mandatory disclosure of pre- and post-mitigation capabilities to approved government bodies, standardized evaluation methods, and minimum transparency requirements for public safety reporting. These ensure that policymakers and regulators can craft targeted safety measures, assess deployment risks, and scrutinize companies' safety claims effectively.
Authors: Atharva Deo, Nicholas Matsumoto, Sun Kim, Peter Wager, Randy G. Tsai, Aaron Denmark, Cherine Yang, Xi Li, Jay Moran, Miguel Hernandez, Andrew J. Hung
Abstract: We present an AI based approach to automate the End-to-end Assessment of Suturing Expertise (EASE), a suturing skills assessment tool that comprehensively defines criteria around relevant sub-skills.1 While EASE provides granular skills assessment related to suturing to provide trainees with an objective evaluation of their aptitude along with actionable insights, the scoring process is currently performed by human evaluators, which is time and resource consuming. The AI based approach solves this by enabling real-time score prediction with minimal resources during model inference. This enables the possibility of real-time feedback to the surgeons/trainees, potentially accelerating the learning process for the suturing task and mitigating critical errors during the surgery, improving patient outcomes. In this study, we focus on the following 7 EASE domains that come under 3 suturing phases: 1) Needle Handling: Number of Repositions, Needle Hold Depth, Needle Hold Ratio, and Needle Hold Angle; 2) Needle Driving: Driving Smoothness, and Wrist Rotation; 3) Needle Withdrawal: Wrist Rotation.
Authors: Michal Prusik, Bartosz Frej, Michal W. Przewozniczek
Abstract: Statistical Linkage Learning (SLL) is a part of many state-of-the-art optimizers. The purpose of SLL is to discover variable interdependencies. It has been shown that the effectiveness of SLL-using optimizers is highly dependent on the quality of SLL-based problem decomposition. Thus, understanding what kind of problems are hard or easy to decompose by SLL is important for practice. In this work, we analytically estimate the size of a population sufficient for obtaining a perfect decomposition in case of concatenations of certain unitation-based functions. The experimental study confirms the accuracy of the proposed estimate. Finally, using the proposed estimate, we identify those problem types that may be considered hard for SLL-using optimizers.
Authors: Qian Chen, Mohamed Elrefaie, Angela Dai, Faez Ahmed
Abstract: Computational Fluid Dynamics (CFD) simulations are essential in product design, providing insights into fluid behavior around complex geometries in aerospace and automotive applications. However, high-fidelity CFD simulations are computationally expensive, making rapid design iterations challenging. To address this, we propose TripNet, Triplane CFD Network, a machine learning-based framework leveraging triplane representations to predict the outcomes of large-scale, high-fidelity CFD simulations with significantly reduced computation cost. Our method encodes 3D geometry into compact yet information-rich triplane features, maintaining full geometry fidelity and enabling accurate aerodynamic predictions. Unlike graph- and point cloud-based models, which are inherently discrete and provide solutions only at the mesh nodes, TripNet allows the solution to be queried at any point in the 3D space. Validated on high-fidelity DrivAerNet and DrivAerNet++ car aerodynamics datasets, TripNet achieves state-of-the-art performance in drag coefficient prediction, surface field estimation, and full 3D flow field simulations of industry-standard car designs. By utilizing a shared triplane backbone across multiple tasks, our approach offers a scalable, accurate, and efficient alternative to traditional CFD solvers.
Authors: Hugh Dance, Pierre Glaser, Peter Orbanz, Ryan Adams
Abstract: With the advent of automatic vectorization tools (e.g., JAX's $\texttt{vmap}$), writing multi-chain MCMC algorithms is often now as simple as invoking those tools on single-chain code. Whilst convenient, for various MCMC algorithms this results in a synchronization problem -- loosely speaking, at each iteration all chains running in parallel must wait until the last chain has finished drawing its sample. In this work, we show how to design single-chain MCMC algorithms in a way that avoids synchronization overheads when vectorizing with tools like $\texttt{vmap}$ by using the framework of finite state machines (FSMs). Using a simplified model, we derive an exact theoretical form of the obtainable speed-ups using our approach, and use it to make principled recommendations for optimal algorithm design. We implement several popular MCMC algorithms as FSMs, including Elliptical Slice Sampling, HMC-NUTS, and Delayed Rejection, demonstrating speed-ups of up to an order of magnitude in experiments.
Authors: Jiaheng Liu, Dawei Zhu, Zhiqi Bai, Yancheng He, Huanxuan Liao, Haoran Que, Zekun Wang, Chenchen Zhang, Ge Zhang, Jiebin Zhang, Yuanxing Zhang, Zhuo Chen, Hangyu Guo, Shilong Li, Ziqiang Liu, Yong Shan, Yifan Song, Jiayi Tian, Wenhao Wu, Zhejian Zhou, Ruijie Zhu, Junlan Feng, Yang Gao, Shizhu He, Zhoujun Li, Tianyu Liu, Fanyu Meng, Wenbo Su, Yingshui Tan, Zili Wang, Jian Yang, Wei Ye, Bo Zheng, Wangchunshu Zhou, Wenhao Huang, Sujian Li, Zhaoxiang Zhang
Abstract: Efficient processing of long contexts has been a persistent pursuit in Natural Language Processing. With the growing number of long documents, dialogues, and other textual data, it is important to develop Long Context Language Models (LCLMs) that can process and analyze extensive inputs in an effective and efficient way. In this paper, we present a comprehensive survey on recent advances in long-context modeling for large language models. Our survey is structured around three key aspects: how to obtain effective and efficient LCLMs, how to train and deploy LCLMs efficiently, and how to evaluate and analyze LCLMs comprehensively. For the first aspect, we discuss data strategies, architectural designs, and workflow approaches oriented with long context processing. For the second aspect, we provide a detailed examination of the infrastructure required for LCLM training and inference. For the third aspect, we present evaluation paradigms for long-context comprehension and long-form generation, as well as behavioral analysis and mechanism interpretability of LCLMs. Beyond these three key aspects, we thoroughly explore the diverse application scenarios where existing LCLMs have been deployed and outline promising future development directions. This survey provides an up-to-date review of the literature on long-context LLMs, which we wish to serve as a valuable resource for both researchers and engineers. An associated GitHub repository collecting the latest papers and repos is available at: \href{https://github.com/LCLM-Horizon/A-Comprehensive-Survey-For-Long-Context-Language-Modeling}{\color[RGB]{175,36,67}{LCLM-Horizon}}.
URLs: https://github.com/LCLM-Horizon/A-Comprehensive-Survey-For-Long-Context-Language-Modeling
Authors: Madhavendra Thakur, Jason Hausenloy
Abstract: Synthetic data, or data generated by machine learning models, is increasingly emerging as a solution to the data access problem. However, its use introduces significant governance and accountability challenges, and potentially debases existing governance paradigms, such as compute and data governance. In this paper, we identify 3 key governance and accountability challenges that synthetic data poses - it can enable the increased emergence of malicious actors, spontaneous biases and value drift. We thus craft 3 technical mechanisms to address these specific challenges, finding applications for synthetic data towards adversarial training, bias mitigation and value reinforcement. These could not only counteract the risks of synthetic data, but serve as critical levers for governance of the frontier in the future.
Authors: Boyue Caroline Hu, Divya Gopinath, Corina S. Pasareanu, Nina Narodytska, Ravi Mangal, Susmit Jha
Abstract: Debugging of Deep Neural Networks (DNNs), particularly vision models, is very challenging due to the complex and opaque decision-making processes in these networks. In this paper, we explore multi-modal Vision-Language Models (VLMs), such as CLIP, to automatically interpret the opaque representation space of vision models using natural language. This in turn, enables a semantic analysis of model behavior using human-understandable concepts, without requiring costly human annotations. Key to our approach is the notion of semantic heatmap, that succinctly captures the statistical properties of DNNs in terms of the concepts discovered with the VLM and that are computed off-line using a held-out data set. We show the utility of semantic heatmaps for fault localization -- an essential step in debugging -- in vision models. Our proposed technique helps localize the fault in the network (encoder vs head) and also highlights the responsible high-level concepts, by leveraging novel differential heatmaps, which summarize the semantic differences between the correct and incorrect behaviour of the analyzed DNN. We further propose a lightweight runtime analysis to detect and filter-out defects at runtime, thus improving the reliability of the analyzed DNNs. The runtime analysis works by measuring and comparing the similarity between the heatmap computed for a new (unseen) input and the heatmaps computed a-priori for correct vs incorrect DNN behavior. We consider two types of defects: misclassifications and vulnerabilities to adversarial attacks. We demonstrate the debugging and runtime analysis on a case study involving a complex ResNet-based classifier trained on the RIVAL10 dataset.
Authors: Veysel Kocaman, Yigit Gul, M. Aytug Kaya, Hasham Ul Haq, Mehmet Butgul, Cabir Celik, David Talby
Abstract: Assertion status detection is a critical yet often overlooked component of clinical NLP, essential for accurately attributing extracted medical facts. Past studies have narrowly focused on negation detection, leading to underperforming commercial solutions such as AWS Medical Comprehend, Azure AI Text Analytics, and GPT-4o due to their limited domain adaptation. To address this gap, we developed state-of-the-art assertion detection models, including fine-tuned LLMs, transformer-based classifiers, few-shot classifiers, and deep learning (DL) approaches. We evaluated these models against cloud-based commercial API solutions, the legacy rule-based NegEx approach, and GPT-4o. Our fine-tuned LLM achieves the highest overall accuracy (0.962), outperforming GPT-4o (0.901) and commercial APIs by a notable margin, particularly excelling in Present (+4.2%), Absent (+8.4%), and Hypothetical (+23.4%) assertions. Our DL-based models surpass commercial solutions in Conditional (+5.3%) and Associated-with-Someone-Else (+10.1%) categories, while the few-shot classifier offers a lightweight yet highly competitive alternative (0.929), making it ideal for resource-constrained environments. Integrated within Spark NLP, our models consistently outperform black-box commercial solutions while enabling scalable inference and seamless integration with medical NER, Relation Extraction, and Terminology Resolution. These results reinforce the importance of domain-adapted, transparent, and customizable clinical NLP solutions over general-purpose LLMs and proprietary APIs.
Authors: Michael D. White, Michael D. Atkinson, Adam J. Plowman, Pratheek Shanthraj
Abstract: Microstructure quantification is an important step towards establishing structure-property relationships in materials. Machine learning-based image processing methods have been shown to outperform conventional image processing techniques and are increasingly applied to microstructure quantification tasks. In this work, we present a 3D variational autoencoder (VAE) for encoding microstructure volume elements (VEs) comprising voxelated crystallographic orientation data. Crystal symmetries in the orientation space are accounted for by mapping to the crystallographic fundamental zone as a preprocessing step, which allows for a continuous loss function to be used and improves the training convergence rate. The VAE is then used to encode a training set of VEs with an equiaxed polycrystalline microstructure with random texture. Accurate reconstructions are achieved with a relative average misorientation error of 9x10-3 on the test dataset, for a continuous latent space with dimension 256. We show that the model generalises well to microstructures with textures, grain sizes and aspect ratios outside the training distribution. Structure-property relationships are explored through using the training set of VEs as initial configurations in various crystal plasticity (CP) simulations. Microstructural fingerprints extracted from the VAE, which parameterise the VEs in a low-dimensional latent space, are stored alongside the volume-averaged stress response, at each strain increment, to uniaxial tensile deformation from CP simulations. This is then used to train a fully connected neural network mapping the input fingerprint to the resulting stress response, which acts as a surrogate model for the CP simulation. The fingerprint-based surrogate model is shown to accurately predict the microstructural dependence in the CP stress response, with a relative mean-squared error of 8.9x10-4 on unseen test data.
Authors: Yi-Meng Huang, Zi-Ran Zhao, Shun-Cai Zhao
Abstract: In machine learning (ML), the risk of recursive strategies overfitting historical data has driven the development of convolutional neural networks (CNNs) in simulating quantum dissipative dynamics. In this work, we propose an efficient CNNs scheme incorporating novel redundant time-functions to predict 100 picosecond (ps) excitation energy transfer (EET) in Fenna-Matthews-Olson (FMO) complexes, in which the original time $t$ is normalized by mapping it to the [0, 1] range, allowing different functions focus on distinct time intervals, thereby effectively capturing the multi-timescale characteristics of EET dynamics. This method simplifies optimization and enhances learning efficiency, and demonstrate the superior accuracy, robustness, and efficiency of our approach in predicting quantum dissipative dynamics.
Authors: Saray Bakker, Rodrigo P\'erez-Dattari, Cosimo Della Santina, Wendelin B\"ohmer, Javier Alonso-Mora
Abstract: Using the language of dynamical systems, Imitation learning (IL) provides an intuitive and effective way of teaching stable task-space motions to robots with goal convergence. Yet, IL techniques are affected by serious limitations when it comes to ensuring safety and fulfillment of physical constraints. With this work, we solve this challenge via TamedPUMA, an IL algorithm augmented with a recent development in motion generation called geometric fabrics. As both the IL policy and geometric fabrics describe motions as artificial second-order dynamical systems, we propose two variations where IL provides a navigation policy for geometric fabrics. The result is a stable imitation learning strategy within which we can seamlessly blend geometrical constraints like collision avoidance and joint limits. Beyond providing a theoretical analysis, we demonstrate TamedPUMA with simulated and real-world tasks, including a 7-DoF manipulator.
Authors: Paolo Frazzetto, Muhammad Uzair Ul Haq, Flavia Fabris, Alessandro Sperduti
Abstract: The recruitment process is undergoing a significant transformation with the increasing use of machine learning and natural language processing techniques. While previous studies have focused on automating candidate selection, the role of multiple vacancies in this process remains understudied. This paper addresses this gap by proposing a novel pipeline that leverages Large Language Models and graph similarity measures to suggest ideal candidates for specific job openings. Our approach represents candidate profiles as multimodal embeddings, enabling the capture of nuanced relationships between job requirements and candidate attributes. The proposed approach has significant implications for the recruitment industry, enabling companies to streamline their hiring processes and identify top talent more efficiently. Our work contributes to the growing body of research on the application of machine learning in human resources, highlighting the potential of LLMs and graph-based methods in revolutionizing the recruitment landscape.
Authors: Soumen Kumar Mondal, Sayambhu Sen, Abhishek Singhania, Preethi Jyothi
Abstract: Multilingual large language models (LLMs) aim towards robust natural language understanding across diverse languages, yet their performance significantly degrades on low-resource languages. This work explores whether existing techniques to identify language-specific neurons can be leveraged to enhance cross-lingual task performance of lowresource languages. We conduct detailed experiments covering existing language-specific neuron identification techniques (such as Language Activation Probability Entropy and activation probability-based thresholding) and neuron-specific LoRA fine-tuning with models like Llama 3.1 and Mistral Nemo. We find that such neuron-specific interventions are insufficient to yield cross-lingual improvements on downstream tasks (XNLI, XQuAD) in lowresource languages. This study highlights the challenges in achieving cross-lingual generalization and provides critical insights for multilingual LLMs.
Authors: Daniel Kuhse, Harun Teper, Sebastian Buschj\"ager, Chien-Yao Wang, Jian-Jia Chen
Abstract: We introduce AnytimeYOLO, a family of variants of the YOLO architecture that enables anytime object detection. Our AnytimeYOLO networks allow for interruptible inference, i.e., they provide a prediction at any point in time, a property desirable for safety-critical real-time applications. We present structured explorations to modify the YOLO architecture, enabling early termination to obtain intermediate results. We focus on providing fine-grained control through high granularity of available termination points. First, we formalize Anytime Models as a special class of prediction models that offer anytime predictions. Then, we discuss a novel transposed variant of the YOLO architecture, that changes the architecture to enable better early predictions and greater freedom for the order of processing stages. Finally, we propose two optimization algorithms that, given an anytime model, can be used to determine the optimal exit execution order and the optimal subset of early-exits to select for deployment in low-resource environments. We evaluate the anytime performance and trade-offs of design choices, proposing a new anytime quality metric for this purpose. In particular, we also discuss key challenges for anytime inference that currently make its deployment costly.
Authors: Ken Ziyu Liu, Christopher A. Choquette-Choo, Matthew Jagielski, Peter Kairouz, Sanmi Koyejo, Percy Liang, Nicolas Papernot
Abstract: An important question today is whether a given text was used to train a large language model (LLM). A \emph{completion} test is often employed: check if the LLM completes a sufficiently complex text. This, however, requires a ground-truth definition of membership; most commonly, it is defined as a member based on the $n$-gram overlap between the target text and any text in the dataset. In this work, we demonstrate that this $n$-gram based membership definition can be effectively gamed. We study scenarios where sequences are \emph{non-members} for a given $n$ and we find that completion tests still succeed. We find many natural cases of this phenomenon by retraining LLMs from scratch after removing all training samples that were completed; these cases include exact duplicates, near-duplicates, and even short overlaps. They showcase that it is difficult to find a single viable choice of $n$ for membership definitions. Using these insights, we design adversarial datasets that can cause a given target sequence to be completed without containing it, for any reasonable choice of $n$. Our findings highlight the inadequacy of $n$-gram membership, suggesting membership definitions fail to account for auxiliary information available to the training algorithm.
Authors: Licong Lin, Song Mei
Abstract: Contrastive learning -- a modern approach to extract useful representations from unlabeled data by training models to distinguish similar samples from dissimilar ones -- has driven significant progress in foundation models. In this work, we develop a new theoretical framework for analyzing data augmentation-based contrastive learning, with a focus on SimCLR as a representative example. Our approach is based on the concept of \emph{approximate sufficient statistics}, which we extend beyond its original definition in \cite{oko2025statistical} for contrastive language-image pretraining (CLIP) using KL-divergence. We generalize it to equivalent forms and general f-divergences, and show that minimizing SimCLR and other contrastive losses yields encoders that are approximately sufficient. Furthermore, we demonstrate that these near-sufficient encoders can be effectively adapted to downstream regression and classification tasks, with performance depending on their sufficiency and the error induced by data augmentation in contrastive learning. Concrete examples in linear regression and topic classification are provided to illustrate the broad applicability of our results.
Authors: T\^am Johan Nguy\^en, Darrick Lee, Bernadette Jana Stolz
Abstract: The behavior of multivariate dynamical processes is often governed by underlying structural connections that relate the components of the system. For example, brain activity which is often measured via time series is determined by an underlying structural graph, where nodes represent neurons or brain regions and edges represent cortical connectivity. Existing methods for inferring structural connections from observed dynamics, such as correlation-based or spectral techniques, may fail to fully capture complex relationships in high-dimensional time series in an interpretable way. Here, we propose the use of path signatures a mathematical framework that encodes geometric and temporal properties of continuous paths to address this problem. Path signatures provide a reparametrization-invariant characterization of dynamical data and, in particular, can be used to compute the lead matrix which reveals lead-lag phenomena. We showcase our approach on time series from coupled oscillators in the Kuramoto model defined on a stochastic block model graph, termed the Kuramoto stochastic block model (KSBM). Using mean-field theory and Gaussian approximations, we analytically derive reduced models of KSBM dynamics in different temporal regimes and theoretically characterize the lead matrix in these settings. Leveraging these insights, we propose a novel signature-based community detection algorithm, achieving exact recovery of structural communities from observed time series in multiple KSBM instances. Our results demonstrate that path signatures provide a novel perspective on analyzing complex neural data and other high-dimensional systems, explicitly exploiting temporal functional relationships to infer underlying structure.
Authors: Eric Lei, Hamed Hassani, Shirin Saeedi Bidokhti
Abstract: Recent efforts in neural compression have focused on the rate-distortion-perception (RDP) tradeoff, where the perception constraint ensures the source and reconstruction distributions are close in terms of a statistical divergence. Theoretical work on RDP describes interesting properties of RDP-optimal compressors without providing constructive and low complexity solutions. While classical rate distortion theory shows that optimal compressors should efficiently pack the space, RDP theory additionally shows that infinite randomness shared between the encoder and decoder may be necessary for RDP optimality. In this paper, we propose neural compressors that are low complexity and benefit from high packing efficiency through lattice coding and shared randomness through shared dithering over the lattice cells. For two important settings, namely infinite shared and zero shared randomness, we analyze the rate, distortion, and perception achieved by our proposed neural compressors and further show optimality in the presence of infinite shared randomness. Experimentally, we investigate the roles these two components of our design, lattice coding and randomness, play in the performance of neural compressors on synthetic and real-world data. We observe that performance improves with more shared randomness and better lattice packing.
Authors: Vishwesh Ramanathan, Tony Xu, Pushpak Pati, Faruk Ahmed, Maged Goubran, Anne L. Martel
Abstract: Prediction tasks in digital pathology are challenging due to the massive size of whole-slide images (WSIs) and the weak nature of training signals. Advances in computing, data availability, and self-supervised learning (SSL) have paved the way for slide-level foundation models (SLFMs) that can improve prediction tasks in low-data regimes. However, working with these models is challenging, with issues such as catastrophic forgetting during fine-tuning and under-utilization of shared information between tasks and modalities. To overcome these two challenges, we propose ModalTune, a novel fine-tuning framework which introduces the Modal Adapter to integrate new modalities without modifying SLFM weights. Additionally, we use large-language models (LLMs) to encode labels as text, capturing semantic relationships and enhancing generalization across multiple tasks and cancer types in a single training recipe. ModalTune achieves state-of-the-art (SOTA) results against both uni-modal and multi-modal models across four cancer types, jointly improving survival and cancer subtype prediction while remaining competitive in pan-cancer settings. Additionally, we show ModalTune is highly generalizable to two out-of-distribution (OOD) datasets. To our knowledge, this is the first unified fine-tuning framework for multi-modal, multi-task, and pan-cancer modeling in digital pathology.
Authors: Sara Bicego, Samuel Gue, Dante Kalise, Nelly Villamizar
Abstract: A computational method for the synthesis of time-optimal feedback control laws for linear nilpotent systems is proposed. The method is based on the use of the bang-bang theorem, which leads to a characterization of the time-optimal trajectory as a parameter-dependent polynomial system for the control switching sequence. A deflated Newton's method is then applied to exhaust all the real roots of the polynomial system. The root-finding procedure is informed by the Hermite quadratic form, which provides a sharp estimate on the number of real roots to be found. In the second part of the paper, the polynomial systems are sampled and solved to generate a synthetic dataset for the construction of a time-optimal deep neural network -- interpreted as a binary classifier -- via supervised learning. Numerical tests in integrators of increasing dimension assess the accuracy, robustness, and real-time-control capabilities of the approximate control law.
Authors: Alhasan Abdellatif, Hannah P. Menke, Julien Maes, Ahmed H. Elsheikh, Florian Doster
Abstract: Accurately capturing the complex interaction between CO2 and water in porous media at the pore scale is essential for various geoscience applications, including carbon capture and storage (CCS). We introduce a comprehensive dataset generated from high-fidelity numerical simulations to capture the intricate interaction between CO2 and water at the pore scale. The dataset consists of 624 2D samples, each of size 512x512 with a resolution of 35 {\mu}m, covering 100 time steps under a constant CO2 injection rate. It includes various levels of heterogeneity, represented by different grain sizes with random variation in spacing, offering a robust testbed for developing predictive models. This dataset provides high-resolution temporal and spatial information crucial for benchmarking machine learning models.
Authors: Akhil Joshi, Sai Teja Erukude, Lior Shamir
Abstract: With the availability of virtually infinite number text documents in digital format, automatic comparison of textual data is essential for extracting meaningful insights that are difficult to identify manually. Many existing tools, including AI and large language models, struggle to provide precise and explainable insights into textual similarities. In many cases they determine the similarity between documents as reflected by the text, rather than the similarities between the subjects being discussed in these documents. This study addresses these limitations by developing an n-gram analysis framework designed to compare documents automatically and uncover explainable similarities. A scoring formula is applied to assigns each of the n-grams with a weight, where the weight is higher when the n-grams are more frequent in both documents, but is penalized when the n-grams are more frequent in the English language. Visualization tools like word clouds enhance the representation of these patterns, providing clearer insights. The findings demonstrate that this framework effectively uncovers similarities between text documents, offering explainable insights that are often difficult to identify manually. This non-parametric approach provides a deterministic solution for identifying similarities across various fields, including biographies, scientific literature, historical texts, and more. Code for the method is publicly available.
Authors: Karol Chlasta, Katarzyna Wisiecka, Krzysztof Krejtz, Izabela Krejtz
Abstract: Well-being is a dynamic construct that evolves over time and fluctuates within individuals, presenting challenges for accurate quantification. Reduced well-being is often linked to depression or anxiety disorders, which are characterised by biases in visual attention towards specific stimuli, such as human faces. This paper introduces a novel approach to AI-assisted screening of affective disorders by analysing visual attention scan paths using convolutional neural networks (CNNs). Data were collected from two studies examining (1) attentional tendencies in individuals diagnosed with major depression and (2) social anxiety. These data were processed using residual CNNs through images generated from eye-gaze patterns. Experimental results, obtained with ResNet architectures, demonstrated an average accuracy of 48% for a three-class system and 62% for a two-class system. Based on these exploratory findings, we propose that this method could be employed in rapid, ecological, and effective mental health screening systems to assess well-being through eye-tracking.
Authors: Adam Atanas, Kai Liu
Abstract: Large language models (LLMs) exhibit impressive capabilities but struggle with reasoning errors due to hallucinations and flawed logic. To investigate their internal representations of reasoning, we introduce ArrangementPuzzle, a novel puzzle dataset with structured solutions and automated stepwise correctness verification. We trained a classifier model on LLM activations on this dataset and found that it achieved over 80% accuracy in predicting reasoning correctness, implying that LLMs internally distinguish between correct and incorrect reasoning steps, with the strongest representations in middle-late Transformer layers. Further analysis reveals that LLMs encode abstract reasoning concepts within the middle activation layers of the transformer architecture, distinguishing logical from semantic equivalence. These findings provide insights into LLM reasoning mechanisms and contribute to improving AI reliability and interpretability, thereby offering the possibility to manipulate and refine LLM reasoning.
Authors: Yuheng Ding, Yusong Wang, Bo Qiang, Jie Yu, Qi Li, Yiran Zhou, Zhenmin Liu
Abstract: Natural products, as metabolites from microorganisms, animals, or plants, exhibit diverse biological activities, making them crucial for drug discovery. Nowadays, existing deep learning methods for natural products research primarily rely on supervised learning approaches designed for specific downstream tasks. However, such one-model-for-a-task paradigm often lacks generalizability and leaves significant room for performance improvement. Additionally, existing molecular characterization methods are not well-suited for the unique tasks associated with natural products. To address these limitations, we have pre-trained a foundation model for natural products based on their unique properties. Our approach employs a novel pretraining strategy that is especially tailored to natural products. By incorporating contrastive learning and masked graph learning objectives, we emphasize evolutional information from molecular scaffolds while capturing side-chain information. Our framework achieves state-of-the-art (SOTA) results in various downstream tasks related to natural product mining and drug discovery. We first compare taxonomy classification with synthesized molecule-focused baselines to demonstrate that current models are inadequate for understanding natural synthesis. Furthermore, by diving into a fine-grained analysis at both the gene and microbial levels, NaFM demonstrates the ability to capture evolutionary information. Eventually, our method is experimented with virtual screening, illustrating informative natural product representations that can lead to more effective identification of potential drug candidates.
Authors: Abtin Pourhadi, Paul Swoboda
Abstract: We present a new state of the art approach for sparse keypoint matching between pairs of images. Our method consists of a fully deep learning based approach combining a visual backbone coupled with a SplineCNN graph neural network for feature processing and a normalized transformer decoder for decoding keypoint correspondences together with the Sinkhorn algorithm. Our method is trained using a contrastive and a hyperspherical loss for better feature representations. We additionally use data augmentation during training. This comparatively simple architecture combining extensive normalization and advanced losses outperforms current state of the art approaches on PascalVOC and SPair-71k datasets by $5.1\%$ and $2.2\%$ respectively compared to BBGM, ASAR, COMMON and GMTR while training for at least $1.7x$ fewer epochs.
Authors: Sharan Maiya, Yinhong Liu, Ramit Debnath, Anna Korhonen
Abstract: Large Language Models (LLMs) are often used as automated judges to evaluate text, but their effectiveness can be hindered by various unintentional biases. We propose using linear classifying probes, trained by leveraging differences between contrasting pairs of prompts, to directly access LLMs' latent knowledge and extract more accurate preferences. Through extensive experiments using models of varying size from four different families and six diverse datasets assessing text quality evaluation and common sense reasoning, we demonstrate that both supervised and unsupervised probing approaches consistently outperform traditional generation-based judgement while maintaining similar computational costs. These probes generalise under domain shifts and can even outperform finetuned evaluators with the same training data size. Our results suggest linear probing offers an accurate, robust and computationally efficient approach for LLM-as-judge tasks while providing interpretable insights into how models encode judgement-relevant knowledge. Our data and code will be openly released in the future.
Authors: Nguyen Phuc Tran, Brigitte Jaumard, Oscar Delgado
Abstract: Advanced Large Language Models (LLMs) have revolutionized various fields, including communication networks, sparking an innovation wave that has led to new applications and services, and significantly enhanced solution schemes. Despite all these impressive developments, most LLMs typically require huge computational resources, resulting in terribly high energy consumption. Thus, this research study proposes an end-to-end pipeline that investigates the trade-off between energy efficiency and model performance for an LLM during fault ticket analysis in communication networks. It further evaluates the pipeline performance using two real-world datasets for the tasks of root cause analysis and response feedback in a communication network. Our results show that an appropriate combination of quantization and pruning techniques is able to reduce energy consumption while significantly improving model performance.
Authors: Morgane Austern, Yuanchuan Guo, Zheng Tracy Ke, Tianle Liu
Abstract: Topic modeling is traditionally applied to word counts without accounting for the context in which words appear. Recent advancements in large language models (LLMs) offer contextualized word embeddings, which capture deeper meaning and relationships between words. We aim to leverage such embeddings to improve topic modeling. We use a pre-trained LLM to convert each document into a sequence of word embeddings. This sequence is then modeled as a Poisson point process, with its intensity measure expressed as a convex combination of $K$ base measures, each corresponding to a topic. To estimate these topics, we propose a flexible algorithm that integrates traditional topic modeling methods, enhanced by net-rounding applied before and kernel smoothing applied after. One advantage of this framework is that it treats the LLM as a black box, requiring no fine-tuning of its parameters. Another advantage is its ability to seamlessly integrate any traditional topic modeling approach as a plug-in module, without the need for modifications Assuming each topic is a $\beta$-H\"{o}lder smooth intensity measure on the embedded space, we establish the rate of convergence of our method. We also provide a minimax lower bound and show that the rate of our method matches with the lower bound when $\beta\leq 1$. Additionally, we apply our method to several datasets, providing evidence that it offers an advantage over traditional topic modeling approaches.
Authors: Matthias Herp, Johannes Brachem, Michael Altenbuchinger, Thomas Kneib
Abstract: Graphical Transformation Models (GTMs) are introduced as a novel approach to effectively model multivariate data with intricate marginals and complex dependency structures non-parametrically, while maintaining interpretability through the identification of varying conditional independencies. GTMs extend multivariate transformation models by replacing the Gaussian copula with a custom-designed multivariate transformation, offering two major advantages. Firstly, GTMs can capture more complex interdependencies using penalized splines, which also provide an efficient regularization scheme. Secondly, we demonstrate how to approximately regularize GTMs using a lasso penalty towards pairwise conditional independencies, akin to Gaussian graphical models. The model's robustness and effectiveness are validated through simulations, showcasing its ability to accurately learn parametric vine copulas and identify conditional independencies. Additionally, the model is applied to a benchmark astrophysics dataset, where the GTM demonstrates favorable performance compared to non-parametric vine copulas in learning complex multivariate distributions.
Authors: Ruijia Zhang, Siliang Zeng, Chenliang Li, Alfredo Garcia, Mingyi Hong
Abstract: The goal of the Inverse reinforcement learning (IRL) task is to identify the underlying reward function and the corresponding optimal policy from a set of expert demonstrations. While most IRL algorithms' theoretical guarantees rely on a linear reward structure, we aim to extend the theoretical understanding of IRL to scenarios where the reward function is parameterized by neural networks. Meanwhile, conventional IRL algorithms usually adopt a nested structure, leading to computational inefficiency, especially in high-dimensional settings. To address this problem, we propose the first two-timescale single-loop IRL algorithm under neural network parameterized reward and provide a non-asymptotic convergence analysis under overparameterization. Although prior optimality results for linear rewards do not apply, we show that our algorithm can identify the globally optimal reward and policy under certain neural network structures. This is the first IRL algorithm with a non-asymptotic convergence guarantee that provably achieves global optimality in neural network settings.
Authors: Alexandru Apostu, Silviu Gheorghe, Andrei H\^iji, Nicolae Cleju, Andrei P\u{a}tra\c{s}cu, Cristian Rusu, Radu Ionescu, Paul Irofti
Abstract: Distributed Denial of Service attacks represent an active cybersecurity research problem. Recent research shifted from static rule-based defenses towards AI-based detection and mitigation. This comprehensive survey covers several key topics. Preeminently, state-of-the-art AI detection methods are discussed. An in-depth taxonomy based on manual expert hierarchies and an AI-generated dendrogram are provided, thus settling DDoS categorization ambiguities. An important discussion on available datasets follows, covering data format options and their role in training AI detection methods together with adversarial training and examples augmentation. Beyond detection, AI based mitigation techniques are surveyed as well. Finally, multiple open research directions are proposed.
Authors: Xiaochen Du, Mengren Liu, Jiayu Peng, Hoje Chun, Alexander Hoffman, Bilge Yildiz, Lin Li, Martin Z. Bazant, Rafael G\'omez-Bombarelli
Abstract: Electrochemical interfaces are crucial in catalysis, energy storage, and corrosion, where their stability and reactivity depend on complex interactions between the electrode, adsorbates, and electrolyte. Predicting stable surface structures remains challenging, as traditional surface Pourbaix diagrams tend to either rely on expert knowledge or costly $\textit{ab initio}$ sampling, and neglect thermodynamic equilibration with the environment. Machine learning (ML) potentials can accelerate static modeling but often overlook dynamic surface transformations. Here, we extend the Virtual Surface Site Relaxation-Monte Carlo (VSSR-MC) method to autonomously sample surface reconstructions modeled under aqueous electrochemical conditions. Through fine-tuning foundational ML force fields, we accurately and efficiently predict surface energetics, recovering known Pt(111) phases and revealing new LaMnO$_\mathrm{3}$(001) surface reconstructions. By explicitly accounting for bulk-electrolyte equilibria, our framework enhances electrochemical stability predictions, offering a scalable approach to understanding and designing materials for electrochemical applications.
Authors: Samira Alkaee Taleghan, Andrew P. Barrett, Walter N. Meier, Farnoush Banaei-Kashani
Abstract: Sea ice plays a critical role in the global climate system and maritime operations, making timely and accurate classification essential. However, traditional manual methods are time-consuming, costly, and have inherent biases. Automating sea ice type classification addresses these challenges by enabling faster, more consistent, and scalable analysis. While both traditional and deep learning approaches have been explored, deep learning models offer a promising direction for improving efficiency and consistency in sea ice classification. However, the absence of a standardized benchmark and comparative study prevents a clear consensus on the best-performing models. To bridge this gap, we introduce \textit{IceBench}, a comprehensive benchmarking framework for sea ice type classification. Our key contributions are threefold: First, we establish the IceBench benchmarking framework which leverages the existing AI4Arctic Sea Ice Challenge dataset as a standardized dataset, incorporates a comprehensive set of evaluation metrics, and includes representative models from the entire spectrum of sea ice type classification methods categorized in two distinct groups, namely, pixel-based classification methods and patch-based classification methods. IceBench is open-source and allows for convenient integration and evaluation of other sea ice type classification methods; hence, facilitating comparative evaluation of new methods and improving reproducibility in the field. Second, we conduct an in-depth comparative study on representative models to assess their strengths and limitations, providing insights for both practitioners and researchers. Third, we leverage IceBench for systematic experiments addressing key research questions on model transferability across seasons (time) and locations (space), data downscaling, and preprocessing strategies.
Authors: Bokai Cao, Xueyuan Lin, Yiyan Qi, Chengjin Xu, Cehao Yang, Jian Guo
Abstract: Market simulator tries to create high-quality synthetic financial data that mimics real-world market dynamics, which is crucial for model development and robust assessment. Despite continuous advancements in simulation methodologies, market fluctuations vary in terms of scale and sources, but existing frameworks often excel in only specific tasks. To address this challenge, we propose Financial Wind Tunnel (FWT), a retrieval-augmented market simulator designed to generate controllable, reasonable, and adaptable market dynamics for model testing. FWT offers a more comprehensive and systematic generative capability across different data frequencies. By leveraging a retrieval method to discover cross-sectional information as the augmented condition, our diffusion-based simulator seamlessly integrates both macro- and micro-level market patterns. Furthermore, our framework allows the simulation to be controlled with wide applicability, including causal generation through "what-if" prompts or unprecedented cross-market trend synthesis. Additionally, we develop an automated optimizer for downstream quantitative models, using stress testing of simulated scenarios via FWT to enhance returns while controlling risks. Experimental results demonstrate that our approach enables the generalizable and reliable market simulation, significantly improve the performance and adaptability of downstream models, particularly in highly complex and volatile market conditions. Our code and data sample is available at https://anonymous.4open.science/r/fwt_-E852
Authors: Jiachen Jiang, Tianyu Ding, Ke Zhang, Jinxin Zhou, Tianyi Chen, Ilya Zharkov, Zhihui Zhu, Luming Liang
Abstract: All-in-one image restoration seeks to recover high-quality images from various types of degradation using a single model, without prior knowledge of the corruption source. However, existing methods often struggle to effectively and efficiently handle multiple degradation types. We present Cat-AIR, a novel \textbf{C}ontent \textbf{A}nd \textbf{T}ask-aware framework for \textbf{A}ll-in-one \textbf{I}mage \textbf{R}estoration. Cat-AIR incorporates an alternating spatial-channel attention mechanism that adaptively balances the local and global information for different tasks. Specifically, we introduce cross-layer channel attentions and cross-feature spatial attentions that allocate computations based on content and task complexity. Furthermore, we propose a smooth learning strategy that allows for seamless adaptation to new restoration tasks while maintaining performance on existing ones. Extensive experiments demonstrate that Cat-AIR achieves state-of-the-art results across a wide range of restoration tasks, requiring fewer FLOPs than previous methods, establishing new benchmarks for efficient all-in-one image restoration.
Authors: Peilin Cao, Ying Geng, Nan Feng, Xiang Zhang, Zhiwen Qi, Zhen Song, Rafiqul Gani
Abstract: As current group contribution (GC) methods are mostly proposed for a wide size-range of molecules, applying them to property prediction of small refrigerant molecules could lead to unacceptable errors. In this sense, for the design of novel refrigerants and refrigeration systems, tailoring GC-based models specifically fitted to refrigerant molecules is of great interest. In this work, databases of potential refrigerant molecules are first collected, focusing on five key properties related to the operational efficiency of refrigeration systems, namely normal boiling point, critical temperature, critical pressure, enthalpy of vaporization, and acentric factor. Based on tailored small-molecule groups, the GC method is combined with machine learning (ML) to model these performance-related properties. Following the development of GC-ML models, their performance is analyzed to highlight the potential group-to-property contributions. Additionally, the refrigerant property databases are extended internally and externally, based on which examples are presented to highlight the significance of the developed models.
Authors: Zheng Wang, Anna Cai, Xinfeng Xie, Zaifeng Pan, Yue Guan, Weiwei Chu, Jie Wang, Shikai Li, Jianyu Huang, Chris Cai, Yuchen Hao, Yufei Ding
Abstract: In this work, we present WLB-LLM, a workLoad-balanced 4D parallelism for large language model training. We first thoroughly analyze the workload imbalance issue in LLM training and identify two primary sources of imbalance at the pipeline parallelism and context parallelism levels. Then, to address the imbalance issue, at the pipeline parallelism level, WLB-LLM incorporates a workload-aware variable-length document packing method to balance the computation and communication workload across micro-batches. Additionally, at the context parallelism level, WLB-LLM introduces a novel fine-grained per-document sharding strategy, ensuring each worker within a context parallelism group has an identical workload. Comprehensive experiments under different model scales demonstrate that WLB-LLM significantly mitigates the workload imbalance during 4D parallelism LLM training and achieves an average speedup of 1.23x when applying WLB-LLM in our internal LLM training framework.
Authors: Moin Uddin Maruf, Sungmin Kim, Zeeshan Ahmad
Abstract: Machine learning interatomic potentials (MLIPs) provide a computationally efficient alternative to quantum mechanical simulations for predicting material properties. Message-passing graph neural networks, commonly used in these MLIPs, rely on local descriptor-based symmetry functions to model atomic interactions. However, such local descriptor-based approaches struggle with systems exhibiting long-range interactions, charge transfer, and compositional heterogeneity. In this work, we develop a new equivariant MLIP incorporating long-range Coulomb interactions through explicit treatment of electronic degrees of freedom, specifically global charge distribution within the system. This is achieved using a charge equilibration scheme based on predicted atomic electronegativities. We systematically evaluate our model across a range of benchmark periodic and non-periodic datasets, demonstrating that it outperforms both short-range equivariant and long-range invariant MLIPs in energy and force predictions. Our approach enables more accurate and efficient simulations of systems with long-range interactions and charge heterogeneity, expanding the applicability of MLIPs in computational materials science.
Authors: Jussi Jokinen, Patrick Ebel, Tuomo Kujala
Abstract: Modern driving involves interactive technologies that can divert attention, increasing the risk of accidents. This paper presents a computational cognitive model that simulates human multitasking while driving. Based on optimal supervisory control theory, the model predicts how multitasking adapts to variations in driving demands, interactive tasks, and automation levels. Unlike previous models, it accounts for context-dependent multitasking across different degrees of driving automation. The model predicts longer in-car glances on straight roads and shorter glances during curves. It also anticipates increased glance durations with driver aids such as lane-centering assistance and their interaction with environmental demands. Validated against two empirical datasets, the model offers insights into driver multitasking amid evolving in-car technologies and automation.
Authors: Dian Xu, Shanshan Wang, Wei Li, Weibing Deng, Feng Gao, Jianmin Shen
Abstract: Modern machine learning, grounded in the Universal Approximation Theorem, has achieved significant success in the study of phase transitions in both equilibrium and non-equilibrium systems. However, identifying the critical points of percolation models using raw configurations remains a challenging and intriguing problem. This paper proposes the use of the Kolmogorov-Arnold Network, which is based on the Kolmogorov-Arnold Representation Theorem, to input raw configurations into a learning model. The results demonstrate that the KAN can indeed predict the critical points of percolation models. Further observation reveals that, apart from models associated with the density of occupied points, KAN is also capable of effectively achieving phase classification for models where the sole alteration pertains to the orientation of spins, resulting in an order parameter that manifests as an external magnetic flux, such as the Ising model.
Authors: Kunal Mukherjee, Zachary Harrison, Saeid Balaneshin
Abstract: Transparency and interpretability are crucial for enhancing customer confidence and user engagement, especially when dealing with black-box Machine Learning (ML)-based recommendation systems. Modern recommendation systems leverage Graph Neural Network (GNN) due to their ability to produce high-quality recommendations in terms of both relevance and diversity. Therefore, the explainability of GNN is especially important for Link Prediction (LP) tasks since recommending relevant items can be viewed as predicting links between users and items. GNN explainability has been a well-studied field, existing methods primarily focus on node or graph-level tasks, leaving a gap in LP explanation techniques. This work introduces Z-REx, a GNN explanation framework designed explicitly for heterogeneous link prediction tasks. Z-REx utilizes structural and attribute perturbation to identify critical sub-structures and important features while reducing the search space by leveraging domain-specific knowledge. In our experimentation, we show the efficacy of Z-REx in generating contextually relevant and human-interpretable explanations for ZiGNN, a GNN-based recommendation engine, using a real-world real-estate dataset from Zillow Group, Inc. We also compare Z-REx to State-of-The-Art (SOTA) GNN explainers to show Z-REx's superiority in producing high-quality human-interpretable explanations.
Authors: Steven Abreu, Sumit Bam Shrestha, Rui-Jie Zhu, Jason Eshraghian
Abstract: Large language models (LLMs) deliver impressive performance but require large amounts of energy. In this work, we present a MatMul-free LLM architecture adapted for Intel's neuromorphic processor, Loihi 2. Our approach leverages Loihi 2's support for low-precision, event-driven computation and stateful processing. Our hardware-aware quantized model on GPU demonstrates that a 370M parameter MatMul-free model can be quantized with no accuracy loss. Based on preliminary results, we report up to 3x higher throughput with 2x less energy, compared to transformer-based LLMs on an edge GPU, with significantly better scaling. Further hardware optimizations will increase throughput and decrease energy consumption. These results show the potential of neuromorphic hardware for efficient inference and pave the way for efficient reasoning models capable of generating complex, long-form text rapidly and cost-effectively.
Authors: Aabid Karim, Abdul Karim, Bhoomika Lohana, Matt Keon, Jaswinder Singh, Abdul Sattar
Abstract: Large Language Models (LLMs) have significantly advanced various fields, particularly coding, mathematical reasoning, and logical problem solving. However, a critical question remains: Do these mathematical reasoning abilities persist when LLMs are presented with culturally adapted math problems? Specifically, how do LLMs perform when faced with math problems embedded in cultural contexts that have no significant representation in main stream web-scale AI training data? To explore this, we generated six synthetic cultural datasets from GSM8K, a widely used benchmark for assessing LLMs' mathematical reasoning skills. While preserving the mathematical logic and numerical values of the original GSM8K test set, we modify cultural elements such as personal names, food items, place names, etc. These culturally adapted datasets provide a more reliable framework for evaluating LLMs' mathematical reasoning under shifting cultural contexts. Our findings reveal that LLMs struggle with math problems when cultural references change, even though the underlying mathematical structure remains constant. Smaller models exhibit greater performance drops compared to larger models. Interestingly, our results also suggest that cultural familiarity can enhance mathematical reasoning. Even models with no explicit mathematical training but exposure to relevant cultural contexts sometimes outperform larger, mathematically proficient models on culturally embedded math problems. This study highlights the impact of cultural context on the mathematical reasoning abilities of LLMs, underscoring the need for more diverse and representative training data to improve robustness in real-world applications. The benchmark data sets and script for reproducing the results are available at https://github.com/akarim23131/Lost_in_Cultural_Translation
URLs: https://github.com/akarim23131/Lost_in_Cultural_Translation
Authors: Raghul Parthipan, Mohit Anand, Hannah M Christensen, Frederic Vitart, Damon J Wischik, Jakob Zscheischler
Abstract: Regularization is a technique to improve generalization of machine learning (ML) models. A common form of regularization in the ML literature is to train on data where similar inputs map to different outputs. This improves generalization by preventing ML models from becoming overconfident in their predictions. This paper shows how using longer timesteps when modelling chaotic Earth systems naturally leads to more of this regularization. We show this in two domains. We explain how using longer model timesteps can improve results and demonstrate that increased regularization is one of the causes. We explain why longer model timesteps lead to improved regularization in these systems and present a procedure to pick the model timestep. We also carry out a benchmarking exercise on ORAS5 ocean reanalysis data to show that a longer model timestep (28 days) than is typically used gives realistic simulations. We suggest that there will be many opportunities to use this type of regularization in Earth system problems because the Earth system is chaotic and the regularization is so easy to implement.
Authors: Hongshu Guo, Sijie Ma, Zechuan Huang, Yuzhi Hu, Zeyuan Ma, Xinglin Zhang, Yue-Jiao Gong
Abstract: Recently, Meta-Black-Box-Optimization (MetaBBO) methods significantly enhance the performance of traditional black-box optimizers through meta-learning flexible and generalizable meta-level policies that excel in dynamic algorithm configuration (DAC) tasks within the low-level optimization, reducing the expertise required to adapt optimizers for novel optimization tasks. Though promising, existing MetaBBO methods heavily rely on human-crafted feature extraction approach to secure learning effectiveness. To address this issue, this paper introduces a novel MetaBBO method that supports automated feature learning during the meta-learning process, termed as RLDE-AFL, which integrates a learnable feature extraction module into a reinforcement learning-based DE method to learn both the feature encoding and meta-level policy. Specifically, we design an attention-based neural network with mantissa-exponent based embedding to transform the solution populations and corresponding objective values during the low-level optimization into expressive landscape features. We further incorporate a comprehensive algorithm configuration space including diverse DE operators into a reinforcement learning-aided DAC paradigm to unleash the behavior diversity and performance of the proposed RLDE-AFL. Extensive benchmark results show that co-training the proposed feature learning module and DAC policy contributes to the superior optimization performance of RLDE-AFL to several advanced DE methods and recent MetaBBO baselines over both synthetic and realistic BBO scenarios. The source codes of RLDE-AFL are available at https://github.com/GMC-DRL/RLDE-AFL.
Authors: Yunfei Yang, Xiaojun Chen, Yuexin Xuan, Zhendong Zhao
Abstract: Model stealing attack is increasingly threatening the confidentiality of machine learning models deployed in the cloud. Recent studies reveal that adversaries can exploit data synthesis techniques to steal machine learning models even in scenarios devoid of real data, leading to data-free model stealing attacks. Existing defenses against such attacks suffer from limitations, including poor effectiveness, insufficient generalization ability, and low comprehensiveness. In response, this paper introduces a novel defense framework named Model-Guardian. Comprising two components, Data-Free Model Stealing Detector (DFMS-Detector) and Deceptive Predictions (DPreds), Model-Guardian is designed to address the shortcomings of current defenses with the help of the artifact properties of synthetic samples and gradient representations of samples. Extensive experiments on seven prevalent data-free model stealing attacks showcase the effectiveness and superior generalization ability of Model-Guardian, outperforming eleven defense methods and establishing a new state-of-the-art performance. Notably, this work pioneers the utilization of various GANs and diffusion models for generating highly realistic query samples in attacks, with Model-Guardian demonstrating accurate detection capabilities.
Authors: Rochana Chaturvedi, Peyman Baghershahi, Sourav Medya, Barbara Di Eugenio
Abstract: Temporal information extraction from unstructured text is essential for contextualizing events and deriving actionable insights, particularly in the medical domain. We address the task of extracting clinical events and their temporal relations using the well-studied I2B2 2012 Temporal Relations Challenge corpus. This task is inherently challenging due to complex clinical language, long documents, and sparse annotations. We introduce GRAPHTREX, a novel method integrating span-based entity-relation extraction, clinical large pre-trained language models (LPLMs), and Heterogeneous Graph Transformers (HGT) to capture local and global dependencies. Our HGT component facilitates information propagation across the document through innovative global landmarks that bridge distant entities. Our method improves the state-of-the-art with 5.5% improvement in the tempeval $F_1$ score over the previous best and up to 8.9% improvement on long-range relations, which presents a formidable challenge. This work not only advances temporal information extraction but also lays the groundwork for improved diagnostic and prognostic models through enhanced temporal reasoning.
Authors: Lorena G Barberia, Belinda Lombard, Norton Trevisan Roman, Tatiane C. M. Sousa
Abstract: Background Advances in machine learning (ML) models have increased the capability of researchers to detect vaccine hesitancy in social media using Natural Language Processing (NLP). A considerable volume of research has identified the persistence of COVID-19 vaccine hesitancy in discourse shared on various social media platforms. Methods Our objective in this study was to conduct a systematic review of research employing sentiment analysis or stance detection to study discourse towards COVID-19 vaccines and vaccination spread on Twitter (officially known as X since 2023). Following registration in the PROSPERO international registry of systematic reviews, we searched papers published from 1 January 2020 to 31 December 2023 that used supervised machine learning to assess COVID-19 vaccine hesitancy through stance detection or sentiment analysis on Twitter. We categorized the studies according to a taxonomy of five dimensions: tweet sample selection approach, self-reported study type, classification typology, annotation codebook definitions, and interpretation of results. We analyzed if studies using stance detection report different hesitancy trends than those using sentiment analysis by examining how COVID-19 vaccine hesitancy is measured, and whether efforts were made to avoid measurement bias. Results Our review found that measurement bias is widely prevalent in studies employing supervised machine learning to analyze sentiment and stance toward COVID-19 vaccines and vaccination. The reporting errors are sufficiently serious that they hinder the generalisability and interpretation of these studies to understanding whether individual opinions communicate reluctance to vaccinate against SARS-CoV-2. Conclusion Improving the reporting of NLP methods is crucial to addressing knowledge gaps in vaccine hesitancy discourse.
Authors: Filip Stefaniuk, Robert \'Slepaczuk
Abstract: The article investigates the usage of Informer architecture for building automated trading strategies for high frequency Bitcoin data. Three strategies using Informer model with different loss functions: Root Mean Squared Error (RMSE), Generalized Mean Absolute Directional Loss (GMADL) and Quantile loss, are proposed and evaluated against the Buy and Hold benchmark and two benchmark strategies based on technical indicators. The evaluation is conducted using data of various frequencies: 5 minute, 15 minute, and 30 minute intervals, over the 6 different periods. Although the Informer-based model with Quantile loss did not outperform the benchmark, two other models achieved better results. The performance of the model using RMSE loss worsens when used with higher frequency data while the model that uses novel GMADL loss function is benefiting from higher frequency data and when trained on 5 minute interval it beat all the other strategies on most of the testing periods. The primary contribution of this study is the application and assessment of the RMSE, GMADL, and Quantile loss functions with the Informer model to forecast future returns, subsequently using these forecasts to develop automated trading strategies. The research provides evidence that employing an Informer model trained with the GMADL loss function can result in superior trading outcomes compared to the buy-and-hold approach.
Authors: Samuel Schmidgall, Michael Moor
Abstract: Progress in scientific discovery is rarely the result of a single "Eureka" moment, but is rather the product of hundreds of scientists incrementally working together toward a common goal. While existing agent workflows are capable of producing research autonomously, they do so in isolation, without the ability to continuously improve upon prior research results. To address these challenges, we introduce AgentRxiv-a framework that lets LLM agent laboratories upload and retrieve reports from a shared preprint server in order to collaborate, share insights, and iteratively build on each other's research. We task agent laboratories to develop new reasoning and prompting techniques and find that agents with access to their prior research achieve higher performance improvements compared to agents operating in isolation (11.4% relative improvement over baseline on MATH-500). We find that the best performing strategy generalizes to benchmarks in other domains (improving on average by 3.3%). Multiple agent laboratories sharing research through AgentRxiv are able to work together towards a common goal, progressing more rapidly than isolated laboratories, achieving higher overall accuracy (13.7% relative improvement over baseline on MATH-500). These findings suggest that autonomous agents may play a role in designing future AI systems alongside humans. We hope that AgentRxiv allows agents to collaborate toward research goals and enables researchers to accelerate discovery.
Authors: Mariia Slobodian, Mykola Kozlenko
Abstract: This paper presents the machine learning approach to the automated classification of a dog's emotional state based on the processing and recognition of audio signals. It offers helpful information for improving human-machine interfaces and developing more precise tools for classifying emotions from acoustic data. The presented model demonstrates an overall accuracy value above 70% for audio signals recorded for one dog.
Authors: Seyyed Danial Nazemi, Mohsen A. Jafari, Andrea Matta
Abstract: Active Inference (AIF) is emerging as a powerful framework for decision-making under uncertainty, yet its potential in engineering applications remains largely unexplored. In this work, we propose a novel dual-layer AIF architecture that addresses both building-level and community-level energy management. By leveraging the free energy principle, each layer adapts to evolving conditions and handles partial observability without extensive sensor information and respecting data privacy. We validate the continuous AIF model against both a perfect optimization baseline and a reinforcement learning-based approach. We also test the community AIF framework under extreme pricing scenarios. The results highlight the model's robustness in handling abrupt changes. This study is the first to show how a distributed AIF works in engineering. It also highlights new opportunities for privacy-preserving and uncertainty-aware control strategies in engineering applications.
Authors: Suman Adhya, Avishek Lahiri, Debarshi Kumar Sanyal, Partha Pratim Das
Abstract: Negative sampling has emerged as an effective technique that enables deep learning models to learn better representations by introducing the paradigm of learn-to-compare. The goal of this approach is to add robustness to deep learning models to learn better representation by comparing the positive samples against the negative ones. Despite its numerous demonstrations in various areas of computer vision and natural language processing, a comprehensive study of the effect of negative sampling in an unsupervised domain like topic modeling has not been well explored. In this paper, we present a comprehensive analysis of the impact of different negative sampling strategies on neural topic models. We compare the performance of several popular neural topic models by incorporating a negative sampling technique in the decoder of variational autoencoder-based neural topic models. Experiments on four publicly available datasets demonstrate that integrating negative sampling into topic models results in significant enhancements across multiple aspects, including improved topic coherence, richer topic diversity, and more accurate document classification. Manual evaluations also indicate that the inclusion of negative sampling into neural topic models enhances the quality of the generated topics. These findings highlight the potential of negative sampling as a valuable tool for advancing the effectiveness of neural topic models.
Authors: Amanpreet Singh Saimbhi
Abstract: The increasing complexity of modern software systems has led to a rise in vulnerabilities that malicious actors can exploit. Traditional methods of vulnerability detection, such as static and dynamic analysis, have limitations in scalability and automation. This paper proposes a novel approach to detecting software vulnerabilities using a combination of code property graphs and machine learning techniques. By leveraging code property graphs, which integrate abstract syntax trees, control flow graphs, and program dependency graphs, we achieve a detailed representation of software code that enhances the accuracy and granularity of vulnerability detection. We introduce various neural network models, including convolutional neural networks adapted for graph data, to process these representations. Our approach provides a scalable and automated solution for vulnerability detection, addressing the shortcomings of existing methods. We also present a newly generated dataset labeled with function-level vulnerability types sourced from open-source repositories. Our contributions include a methodology for transforming software code into code property graphs, the implementation of a convolutional neural network model for graph data, and the creation of a comprehensive dataset for training and evaluation. This work lays the foundation for more effective and efficient vulnerability detection in complex software systems.
Authors: Spyridon Evangelatos, Eleni Veroni, Vasilis Efthymiou, Christos Nikolopoulos, Georgios Th. Papadopoulos, Panagiotis Sarigiannidis
Abstract: Counterfactual explanations have emerged as a prominent method in Explainable Artificial Intelligence (XAI), providing intuitive and actionable insights into Machine Learning model decisions. In contrast to other traditional feature attribution methods that assess the importance of input variables, counterfactual explanations focus on identifying the minimal changes required to alter a model's prediction, offering a ``what-if'' analysis that is close to human reasoning. In the context of XAI, counterfactuals enhance transparency, trustworthiness and fairness, offering explanations that are not just interpretable but directly applicable in the decision-making processes. In this paper, we present a novel framework that integrates perturbation theory and statistical mechanics to generate minimal counterfactual explanations in explainable AI. We employ a local Taylor expansion of a Machine Learning model's predictive function and reformulate the counterfactual search as an energy minimization problem over a complex landscape. In sequence, we model the probability of candidate perturbations leveraging the Boltzmann distribution and use simulated annealing for iterative refinement. Our approach systematically identifies the smallest modifications required to change a model's prediction while maintaining plausibility. Experimental results on benchmark datasets for cybersecurity in Internet of Things environments, demonstrate that our method provides actionable, interpretable counterfactuals and offers deeper insights into model sensitivity and decision boundaries in high-dimensional spaces.
Authors: Alejandra Castillo, Jamie Haddock, Iryna Hartsock, Paulina Hoyos, Lara Kassab, Alona Kryshchenko, Kamila Larripa, Deanna Needell, Shambhavi Suryanarayanan, Karamatou Yacoubou Djima
Abstract: The reconstruction of tensor-valued signals from corrupted measurements, known as tensor regression, has become essential in many multi-modal applications such as hyperspectral image reconstruction and medical imaging. In this work, we address the tensor linear system problem $\mathcal{A} \mathcal{X}=\mathcal{B}$, where $\mathcal{A}$ is a measurement operator, $\mathcal{X}$ is the unknown tensor-valued signal, and $\mathcal{B}$ contains the measurements, possibly corrupted by arbitrary errors. Such corruption is common in large-scale tensor data, where transmission, sensory, or storage errors are rare per instance but likely over the entire dataset and may be arbitrarily large in magnitude. We extend the Kaczmarz method, a popular iterative algorithm for solving large linear systems, to develop a Quantile Tensor Randomized Kaczmarz (QTRK) method robust to large, sparse corruptions in the observations $\mathcal{B}$. This approach combines the tensor Kaczmarz framework with quantile-based statistics, allowing it to mitigate adversarial corruptions and improve convergence reliability. We also propose and discuss the Masked Quantile Randomized Kaczmarz (mQTRK) variant, which selectively applies partial updates to handle corruptions further. We present convergence guarantees, discuss the advantages and disadvantages of our approaches, and demonstrate the effectiveness of our methods through experiments, including an application for video deblurring.
Authors: Kanishka Parankusham, Rodrigue Rizk, KC Santosh
Abstract: Lakota, a critically endangered language of the Sioux people in North America, faces significant challenges due to declining fluency among younger generations. This paper introduces LakotaBERT, the first large language model (LLM) tailored for Lakota, aiming to support language revitalization efforts. Our research has two primary objectives: (1) to create a comprehensive Lakota language corpus and (2) to develop a customized LLM for Lakota. We compiled a diverse corpus of 105K sentences in Lakota, English, and parallel texts from various sources, such as books and websites, emphasizing the cultural significance and historical context of the Lakota language. Utilizing the RoBERTa architecture, we pre-trained our model and conducted comparative evaluations against established models such as RoBERTa, BERT, and multilingual BERT. Initial results demonstrate a masked language modeling accuracy of 51% with a single ground truth assumption, showcasing performance comparable to that of English-based models. We also evaluated the model using additional metrics, such as precision and F1 score, to provide a comprehensive assessment of its capabilities. By integrating AI and linguistic methodologies, we aspire to enhance linguistic diversity and cultural resilience, setting a valuable precedent for leveraging technology in the revitalization of other endangered indigenous languages.
Authors: Mahak Shah, Akaash Vishal Hazarika, Meetu Malhotra, Sachin C. Patil, Joshit Mohanty
Abstract: Sentiment analysis is a field within NLP that has gained importance because it is applied in various areas such as; social media surveillance, customer feedback evaluation and market research. At the same time, distributed systems allow for effective processing of large amounts of data. Therefore, this paper examines how sentiment analysis converges with distributed systems by concentrating on different approaches, challenges and future investigations. Furthermore, we do an extensive experiment where we train sentiment analysis models using both single node configuration and distributed architecture to bring out the benefits and shortcomings of each method in terms of performance and accuracy.
Authors: Akaash Vishal Hazarika, Mahak Shah, Swapnil Patil, Pradyumna Shukla
Abstract: Effective risk management solutions become absolutely crucial when financial markets embrace distributed technology and decentralized financing (DeFi). This study offers a thorough survey and comparative analysis of the integration of artificial intelligence (AI) in risk management for distributed arbitrage systems. We examine several modern caching techniques namely in memory caching, distributed caching, and proxy caching and their functions in enhancing performance in decentralized settings. Through literature review we examine the utilization of AI techniques for alleviating risks related to market volatility, liquidity challenges, operational failures, regulatory compliance, and security threats. This comparison research evaluates various case studies from prominent DeFi technologies, emphasizing critical performance metrics like latency reduction, load balancing, and system resilience. Additionally, we examine the problems and trade offs associated with these technologies, emphasizing their effects on consistency, scalability, and fault tolerance. By meticulously analyzing real world applications, specifically centering on the Aave platform as our principal case study, we illustrate how the purposeful amalgamation of AI with contemporary caching methodologies has revolutionized risk management in distributed arbitrage systems.
Authors: Jiacheng Yao, Wei Shi, Wei Xu, Zhaohui Yang, A. Lee Swindlehurst, Dusit Niyato
Abstract: Over-the-air computation (AirComp) has emerged as an essential approach for enabling communication-efficient federated learning (FL) over wireless networks. Nonetheless, the inherent analog transmission mechanism in AirComp-based FL (AirFL) intensifies challenges posed by potential Byzantine attacks. In this paper, we propose a novel Byzantine-robust FL paradigm for over-the-air transmissions, referred to as federated learning with secure adaptive clustering (FedSAC). FedSAC aims to protect a portion of the devices from attacks through zero trust architecture (ZTA) based Byzantine identification and adaptive device clustering. By conducting a one-step convergence analysis, we theoretically characterize the convergence behavior with different device clustering mechanisms and uneven aggregation weighting factors for each device. Building upon our analytical results, we formulate a joint optimization problem for the clustering and weighting factors in each communication round. To facilitate the targeted optimization, we propose a dynamic Byzantine identification method using historical reputation based on ZTA. Furthermore, we introduce a sequential clustering method, transforming the joint optimization into a weighting optimization problem without sacrificing the optimality. To optimize the weighting, we capitalize on the penalty convex-concave procedure (P-CCP) to obtain a stationary solution. Numerical results substantiate the superiority of the proposed FedSAC over existing methods in terms of both test accuracy and convergence rate.
Authors: Qingyue Long, Can Rong, Huandong Wang, Shaw Rajib, Yong Li
Abstract: In the real world, trajectory data is often sparse and incomplete due to low collection frequencies or limited device coverage. Trajectory recovery aims to recover these missing trajectory points, making the trajectories denser and more complete. However, this task faces two key challenges: 1) The excessive sparsity of individual trajectories makes it difficult to effectively leverage historical information for recovery; 2) Sparse trajectories make it harder to capture complex individual mobility preferences. To address these challenges, we propose a novel method called DiffMove. Firstly, we harness crowd wisdom for trajectory recovery. Specifically, we construct a group tendency graph using the collective trajectories of all users and then integrate the group mobility trends into the location representations via graph embedding. This solves the challenge of sparse trajectories being unable to rely on individual historical trajectories for recovery. Secondly, we capture individual mobility preferences from both historical and current perspectives. Finally, we integrate group mobility tendencies and individual preferences into the spatiotemporal distribution of the trajectory to recover high-quality trajectories. Extensive experiments on two real-world datasets demonstrate that DiffMove outperforms existing state-of-the-art methods. Further analysis validates the robustness of our method.
Authors: Zhidi Lin, Ying Li, Feng Yin, Juan Maro\~nas, Alexandre H. Thi\'ery
Abstract: Gaussian process state-space models (GPSSMs) have emerged as a powerful framework for modeling dynamical systems, offering interpretable uncertainty quantification and inherent regularization. However, existing GPSSMs face significant challenges in handling high-dimensional, non-stationary systems due to computational inefficiencies, limited scalability, and restrictive stationarity assumptions. In this paper, we propose an efficient transformed Gaussian process state-space model (ETGPSSM) to address these limitations. Our approach leverages a single shared Gaussian process (GP) combined with normalizing flows and Bayesian neural networks, enabling efficient modeling of complex, high-dimensional state transitions while preserving scalability. To address the lack of closed-form expressions for the implicit process in the transformed GP, we follow its generative process and introduce an efficient variational inference algorithm, aided by the ensemble Kalman filter (EnKF), to enable computationally tractable learning and inference. Extensive empirical evaluations on synthetic and real-world datasets demonstrate the superior performance of our ETGPSSM in system dynamics learning, high-dimensional state estimation, and time-series forecasting, outperforming existing GPSSMs and neural network-based methods in both accuracy and computational efficiency.
Authors: Zehao Li, Yijie Peng
Abstract: This paper tackles the challenge of parameter calibration in stochastic models, particularly in scenarios where the likelihood function is unavailable in an analytical form. We introduce a gradient-based simulated parameter estimation framework, which employs a multi-time scale stochastic approximation algorithm. This approach effectively addresses the ratio bias that arises in both maximum likelihood estimation and posterior density estimation problems. The proposed algorithm enhances estimation accuracy and significantly reduces computational costs, as demonstrated through extensive numerical experiments. Our work extends the GSPE framework to handle complex models such as hidden Markov models and variational inference-based problems, offering a robust solution for parameter estimation in challenging stochastic environments.
Authors: Xu Han, Yuan Tang, Jinfeng Xu, Xianzhi Li
Abstract: We introduce Monarch Sparse Tuning (MoST), the first reparameterization-based parameter-efficient fine-tuning (PEFT) method tailored for 3D representation learning. Unlike existing adapter-based and prompt-tuning 3D PEFT methods, MoST introduces no additional inference overhead and is compatible with many 3D representation learning backbones. At its core, we present a new family of structured matrices for 3D point clouds, Point Monarch, which can capture local geometric features of irregular points while offering high expressiveness. MoST reparameterizes the dense update weight matrices as our sparse Point Monarch matrices, significantly reducing parameters while retaining strong performance. Experiments on various backbones show that MoST is simple, effective, and highly generalizable. It captures local features in point clouds, achieving state-of-the-art results on multiple benchmarks, e.g., 97.5% acc. on ScanObjectNN (PB_50_RS) and 96.2% on ModelNet40 classification, while it can also combine with other matrix decompositions (e.g., Low-rank, Kronecker) to further reduce parameters.
Authors: Hankyul Kang, Gregor Seifer, Donghyun Lee, Jongbin Ryu
Abstract: According to the forgetting curve theory, we can enhance memory retention by learning extensive data and taking adequate rest. This means that in order to effectively retain new knowledge, it is essential to learn it thoroughly and ensure sufficient rest so that our brain can memorize without forgetting. The main takeaway from this theory is that learning extensive data at once necessitates sufficient rest before learning the same data again. This aspect of human long-term memory retention can be effectively utilized to address the continual learning of neural networks. Retaining new knowledge for a long period of time without catastrophic forgetting is the critical problem of continual learning. Therefore, based on Ebbinghaus' theory, we introduce the view-batch model that adjusts the learning schedules to optimize the recall interval between retraining the same samples. The proposed view-batch model allows the network to get enough rest to learn extensive knowledge from the same samples with a recall interval of sufficient length. To this end, we specifically present two approaches: 1) a replay method that guarantees the optimal recall interval, and 2) a self-supervised learning that acquires extensive knowledge from a single training sample at a time. We empirically show that these approaches of our method are aligned with the forgetting curve theory, which can enhance long-term memory. In our experiments, we also demonstrate that our method significantly improves many state-of-the-art continual learning methods in various protocols and scenarios. We open-source this project at https://github.com/hankyul2/ViewBatchModel.
Authors: Xusheng Cao, Haori Lu, Linlan Huang, Fei Yang, Xialei Liu, Ming-Ming Cheng
Abstract: Continual learning in computer vision faces the critical challenge of catastrophic forgetting, where models struggle to retain prior knowledge while adapting to new tasks. Although recent studies have attempted to leverage the generalization capabilities of pre-trained models to mitigate overfitting on current tasks, models still tend to forget details of previously learned categories as tasks progress, leading to misclassification. To address these limitations, we introduce a novel Knowledge Graph Enhanced Generative Multi-modal model (KG-GMM) that builds an evolving knowledge graph throughout the learning process. Our approach utilizes relationships within the knowledge graph to augment the class labels and assigns different relations to similar categories to enhance model differentiation. During testing, we propose a Knowledge Graph Augmented Inference method that locates specific categories by analyzing relationships within the generated text, thereby reducing the loss of detailed information about old classes when learning new knowledge and alleviating forgetting. Experiments demonstrate that our method effectively leverages relational information to help the model correct mispredictions, achieving state-of-the-art results in both conventional CIL and few-shot CIL settings, confirming the efficacy of knowledge graphs at preserving knowledge in the continual learning scenarios.
Authors: Xu Fan, Yuetan Lin, Bing Gong, Hao Li
Abstract: Air pollution has become a major threat to human health, making accurate forecasting crucial for pollution control. Traditional physics-based models forecast global air pollution by coupling meteorology and pollution processes, using either online or offline methods depending on whether fully integrated with meteorological models and run simultaneously. However, the high computational demands of both methods severely limit real-time prediction efficiency. Existing deep learning (DL) solutions employ online coupling strategies for global air pollution forecasting, which finetune pollution forecasting based on pretrained atmospheric models, requiring substantial training resources. This study pioneers a DL-based offline coupling framework that utilizes bilinear pooling to achieve offline coupling between meteorological fields and pollutants. The proposed model requires only 13% of the parameters of DL-based online coupling models while achieving competitive performance. Compared with the state-of-the-art global air pollution forecasting model CAMS, our approach demonstrates superiority in 63% variables across all forecast time steps and 85% variables in predictions exceeding 48 hours. This work pioneers experimental validation of the effectiveness of meteorological fields in DL-based global air pollution forecasting, demonstrating that offline coupling meteorological fields with pollutants can achieve a 15% relative reduction in RMSE across all pollution variables. The research establishes a new paradigm for real-time global air pollution warning systems and delivers critical technical support for developing more efficient and comprehensive AI-powered global atmospheric forecasting frameworks.
Authors: Junsong Li, Jie Zhou, Yutao Yang, Bihao Zhan, Qianjun Pan, Yuyang Ding, Qin Chen, Jiang Bo, Xin Lin, Liang He
Abstract: Automatic math correction aims to check students' solutions to mathematical problems via artificial intelligence technologies. Most existing studies focus on judging the final answer at the problem level, while they ignore detailed feedback on each step in a math problem-solving process, which requires abilities of semantic understanding and reasoning. In this paper, we propose a reinforcement learning (RL)-based method to boost large language model (LLM) for step-level automatic math correction, named StepAMC. Particularly, we convert the step-level automatic math correction within the text classification task into an RL problem to enhance the reasoning capabilities of LLMs. Then, we design a space-constrained policy network to improve the stability of RL. Then, we introduce a fine-grained reward network to convert the binary human feedback into a continuous value. We conduct extensive experiments over two benchmark datasets and the results show that our model outperforms the eleven strong baselines.
Authors: Yunhong Lu, Qichao Wang, Hengyuan Cao, Xierui Wang, Xiaoyin Xu, Min Zhang
Abstract: Without using explicit reward, direct preference optimization (DPO) employs paired human preference data to fine-tune generative models, a method that has garnered considerable attention in large language models (LLMs). However, exploration of aligning text-to-image (T2I) diffusion models with human preferences remains limited. In comparison to supervised fine-tuning, existing methods that align diffusion model suffer from low training efficiency and subpar generation quality due to the long Markov chain process and the intractability of the reverse process. To address these limitations, we introduce DDIM-InPO, an efficient method for direct preference alignment of diffusion models. Our approach conceptualizes diffusion model as a single-step generative model, allowing us to fine-tune the outputs of specific latent variables selectively. In order to accomplish this objective, we first assign implicit rewards to any latent variable directly via a reparameterization technique. Then we construct an Inversion technique to estimate appropriate latent variables for preference optimization. This modification process enables the diffusion model to only fine-tune the outputs of latent variables that have a strong correlation with the preference dataset. Experimental results indicate that our DDIM-InPO achieves state-of-the-art performance with just 400 steps of fine-tuning, surpassing all preference aligning baselines for T2I diffusion models in human preference evaluation tasks.
Authors: Dawit Ketema Gete, Bedru Yimam Ahamed, Tadesse Destaw Belay, Yohannes Ayana Ejigu, Sukairaj Hafiz Imam, Alemu Belay Tessema, Mohammed Oumer Adem, Tadesse Amare Belay, Robert Geislinger, Umma Aliyu Musa, Martin Semmann, Shamsuddeen Hassan Muhammad, Henning Schreiber, Seid Muhie Yimam
Abstract: This work explores fine-tuning OpenAI's Whisper automatic speech recognition (ASR) model for Amharic, a low-resource language, to improve transcription accuracy. While the foundational Whisper model struggles with Amharic due to limited representation in its training data, we fine-tune it using datasets like Mozilla Common Voice, FLEURS, and the BDU-speech dataset. The best-performing model, Whispersmall-am, significantly improves when finetuned on a mix of existing FLEURS data and new, unseen Amharic datasets. Training solely on new data leads to poor performance, but combining it with FLEURS data reinforces the model, enabling better specialization in Amharic. We also demonstrate that normalizing Amharic homophones significantly enhances Word Error Rate (WER) and Bilingual Evaluation Understudy (BLEU) scores. This study underscores the importance of fine-tuning strategies and dataset composition for improving ASR in low-resource languages, providing insights for future Amharic speech recognition research.
Authors: Hao-Yuan Chen, Cheng-Pong Huang, Jui-Ming Yao
Abstract: The emergence of large language models and their applications as AI agents have significantly advanced state-of-the-art code generation benchmarks, transforming modern software engineering tasks. However, even with test-time computed reasoning models, these systems still struggle with complex software engineering challenges. This work introduces CURA, a code understanding and reasoning agent system enhanced with verbal process supervision (VPS), achieving a 3.65\% improvement over baseline models on challenging benchmarks like BigCodeBench. Furthermore, CURA, when paired with the o3-mini model and VPS techniques, attains state-of-the-art performance. This work represents a step forward in integrating reasoning-driven architectures with LLM-based code generation, enabling agentic reasoning for language models to solve complex software engineering tasks.
Authors: Yujing Liu, Zhixin Liu, Lei Guo
Abstract: Mixed linear regression (MLR) has attracted increasing attention because of its great theoretical and practical importance in capturing nonlinear relationships by utilizing a mixture of linear regression sub-models. Although considerable efforts have been devoted to the learning problem of such systems, i.e., estimating data labels and identifying model parameters, most existing investigations employ the offline algorithm, impose the strict independent and identically distributed (i.i.d.) or persistent excitation (PE) conditions on the regressor data, and provide local convergence results only. In this paper, we investigate the recursive estimation and data clustering problems for a class of stochastic MLRs with two components. To address this inherently nonconvex optimization problem, we propose a novel two-step recursive identification algorithm to estimate the true parameters, where the direction vector and the scaling coefficient of the unknown parameters are estimated by the least squares and the expectation-maximization (EM) principles, respectively. Under a general data condition, which is much weaker than the traditional i.i.d. and PE conditions, we establish the global convergence and the convergence rate of the proposed identification algorithm for the first time. Furthermore, we prove that, without any excitation condition on the regressor data, the data clustering performance including the cumulative mis-classification error and the within-cluster error can be optimal asymptotically. Finally, we provide a numerical example to illustrate the performance of the proposed learning algorithm.
Authors: Nijesh Upreti, Vaishak Belle
Abstract: Weak supervision allows machine learning models to learn from limited or noisy labels, but it introduces challenges in interpretability and reliability - particularly in multi-instance partial label learning (MI-PLL), where models must resolve both ambiguous labels and uncertain instance-label mappings. We propose a semantics for neuro-symbolic framework that integrates Inductive Logic Programming (ILP) to improve MI-PLL by providing structured relational constraints that guide learning. Within our semantic characterization, ILP defines a logical hypothesis space for label transitions, clarifies classifier semantics, and establishes interpretable performance standards. This hybrid approach improves robustness, transparency, and accountability in weakly supervised settings, ensuring neural predictions align with domain knowledge. By embedding weak supervision into a logical framework, we enhance both interpretability and learning, making weak supervision more suitable for real-world, high-stakes applications.
Authors: Octi Zhang, Quanquan Peng, Rosario Scalise, Bryon Boots
Abstract: Developing robotic agents that can perform well in diverse environments while showing a variety of behaviors is a key challenge in AI and robotics. Traditional reinforcement learning (RL) methods often create agents that specialize in narrow tasks, limiting their adaptability and diversity. To overcome this, we propose a preliminary, evolution-inspired framework that includes a reproduction module, similar to natural species reproduction, balancing diversity and specialization. By integrating RL, imitation learning (IL), and a coevolutionary agent-terrain curriculum, our system evolves agents continuously through complex tasks. This approach promotes adaptability, inheritance of useful traits, and continual learning. Agents not only refine inherited skills but also surpass their predecessors. Our initial experiments show that this method improves exploration efficiency and supports open-ended learning, offering a scalable solution where sparse reward coupled with diverse terrain environments induces a multi-task setting.
Authors: Ashenafi Zebene Woldaregay, J{\o}rgen Aarmo Lund, Phuong Dinh Ngo, Mariyam Tayefi, Joel Burman, Stine Hansen, Martin Hylleholt Sillesen, Hercules Dalianis, Robert Jenssen, Lindsetmo Rolf Ole, Karl {\O}yvind Mikalsen
Abstract: Background: Clinical natural language processing (NLP) refers to the use of computational methods for extracting, processing, and analyzing unstructured clinical text data, and holds a huge potential to transform healthcare in various clinical tasks. Objective: The study aims to perform a systematic review to comprehensively assess and analyze the state-of-the-art NLP methods for the mainland Scandinavian clinical text. Method: A literature search was conducted in various online databases including PubMed, ScienceDirect, Google Scholar, ACM digital library, and IEEE Xplore between December 2022 and February 2024. Further, relevant references to the included articles were also used to solidify our search. The final pool includes articles that conducted clinical NLP in the mainland Scandinavian languages and were published in English between 2010 and 2024. Results: Out of the 113 articles, 18% (n=21) focus on Norwegian clinical text, 64% (n=72) on Swedish, 10% (n=11) on Danish, and 8% (n=9) focus on more than one language. Generally, the review identified positive developments across the region despite some observable gaps and disparities between the languages. There are substantial disparities in the level of adoption of transformer-based models. In essential tasks such as de-identification, there is significantly less research activity focusing on Norwegian and Danish compared to Swedish text. Further, the review identified a low level of sharing resources such as data, experimentation code, pre-trained models, and rate of adaptation and transfer learning in the region. Conclusion: The review presented a comprehensive assessment of the state-of-the-art Clinical NLP for electronic health records (EHR) text in mainland Scandinavian languages and, highlighted the potential barriers and challenges that hinder the rapid advancement of the field in the region.
Authors: Nariman Naderi, Seyed Amir Ahmad Safavi-Naini, Thomas Savage, Zahra Atf, Peter Lewis, Girish Nadkarni, Ali Soroush
Abstract: This study evaluated self-reported response certainty across several large language models (GPT, Claude, Llama, Phi, Mistral, Gemini, Gemma, and Qwen) using 300 gastroenterology board-style questions. The highest-performing models (GPT-o1 preview, GPT-4o, and Claude-3.5-Sonnet) achieved Brier scores of 0.15-0.2 and AUROC of 0.6. Although newer models demonstrated improved performance, all exhibited a consistent tendency towards overconfidence. Uncertainty estimation presents a significant challenge to the safe use of LLMs in healthcare. Keywords: Large Language Models; Confidence Elicitation; Artificial Intelligence; Gastroenterology; Uncertainty Quantification
Authors: Abhijith M S, Sandra S
Abstract: The study proposes a data-driven model which combines the Dynamic Mode Decomposition with multi-linear interpolation to predict the thermal fields of nanofluid flows at unseen Reynolds numbers (Re) and particle volume concentrations ($\epsilon$). The flow, considered for the study, is laminar and incompressible. The study employs an in-house Fortran-based solver to predict the thermal fields of Al$_2$O$_3$-water nanofluid flow through a two-dimensional rectangular channel, with the bottom wall subjected to a uniform heat flux. The performance of two models operating in one- and two-dimensional parametric spaces are investigated. Initially, a DMD with linear interpolation (DMD-LI) based solver is used for prediction of temperature of the nanofluid at any Re $>$ 100. The DMD-LI based model, predicts temperature fields with a maximum percentage difference of just 0.0273\%, in comparison with the CFD-based solver at Re =960, and $\epsilon$ = 1.0\%. The corresponding difference in the average Nusselt numbers is only 0.39\%. Following that a DMD with bi-linear interpolation (DMD-BLI) based solver is used for prediction of temperature of the nanofluid at any Re $>$ 100 and $\epsilon$ $>$ 0.5\%. The performance of two different ways of stacking the data are also examined. When compared to the CFD-based model, the DMD-BLI-based model predicts the temperature fields with a maximum percentage difference of 0.21 \%, at Re = 800 and $\epsilon$ = 1.35\%. And the corresponding percentage difference in the average Nusselt number prediction is only 6.08\%. All the results are reported in detail. Along side the important conclusions, the future scope of the study is also listed.
Authors: Maria Panagiotou, Knut Stroemmen, Lorenzo Brigato, Bastiaan E. de Galan, Stavroula Mougiakakou
Abstract: The growing worldwide incidence of diabetes requires more effective approaches for managing blood glucose levels. Insulin delivery systems have advanced significantly, with artificial intelligence (AI) playing a key role in improving their precision and adaptability. AI algorithms, particularly those based on reinforcement learning, allow for personalised insulin dosing by continuously adapting to an individual's responses. Despite these advancements, challenges such as data privacy, algorithm transparency, and accessibility still need to be addressed. Continued progress and validation in AI-driven insulin delivery systems promise to improve therapy outcomes further, offering people more effective and individualised management of their diabetes. This paper presents an overview of current strategies, key challenges, and future directions.
Authors: Minsu Kim, Seongmin Hong, RyeoWook Ko, Soongyu Choi, Hunjong Lee, Junsoo Kim, Joo-Young Kim, Jongse Park
Abstract: Modern Large Language Model serving system batches multiple requests to achieve high throughput, while batching attention operations is challenging, rendering memory bandwidth a critical bottleneck. The community relies on high-end GPUs with multiple high-bandwidth memory channels. Unfortunately, HBM's high bandwidth often comes at the expense of limited memory capacity, which reduces core utilization and increases costs. Recent advancements enabling longer contexts for LLMs have substantially increased the key-value cache size, further intensifying the pressures on memory capacity. The literature has explored KV cache quantization techniques, which commonly use low bitwidth for most values, selectively using higher bitwidth for outlier values. While this approach helps achieve high accuracy and low bitwidth simultaneously, it comes with the limitation that cost for online outlier detection is excessively high, negating the advantages. We propose Oaken, an acceleration solution that achieves high accuracy and high performance simultaneously through co-designing algorithm and hardware. To effectively find a sweet spot in the accuracy-performance trade-off space of KV cache quantization, Oaken employs an online-offline hybrid approach, setting outlier thresholds offline, which are then used to determine the quantization scale online. To translate the proposed algorithmic technique into tangible performance gains, Oaken also comes with custom quantization engines and memory management units that can be integrated with any LLM accelerators. We built an Oaken accelerator on top of an LLM accelerator, LPU, and conducted a comprehensive evaluation. Our experiments show that for a batch size of 256, Oaken achieves up to 1.58x throughput improvement over NVIDIA A100 GPU, incurring a minimal accuracy loss of only 0.54\% on average, compared to state-of-the-art KV cache quantization techniques.
Authors: Toby St Clere Smithe, Marco Perin
Abstract: We introduce a new compositional framework for generalized variational inference, clarifying the different parts of a model, how they interact, and how they compose. We explain that both exact Bayesian inference and the loss functions typical of variational inference (such as variational free energy and its generalizations) satisfy chain rules akin to that of reverse-mode automatic differentiation, and we advocate for exploiting this to build and optimize models accordingly. To this end, we construct a series of compositional tools: for building models; for constructing their inversions; for attaching local loss functions; and for exposing parameters. Finally, we explain how the resulting parameterized statistical games may be optimized locally, too. We illustrate our framework with a number of classic examples, pointing to new areas of extensibility that are revealed.
Authors: Tomasz R\'o\.za\'nski, Yuan-Sen Ting
Abstract: Neural network-based emulators for the inference of stellar parameters and elemental abundances represent an increasingly popular methodology in modern spectroscopic surveys. However, these approaches are often constrained by their emulation precision and domain transfer capabilities. Greater generalizability has previously been achieved only with significantly larger model architectures, as demonstrated by Transformer-based models in natural language processing. This observation aligns with neural scaling laws, where model performance predictably improves with increased model size, computational resources allocated to model training, and training data volume. In this study, we demonstrate that these scaling laws also apply to Transformer-based spectral emulators in astronomy. Building upon our previous work with TransformerPayne and incorporating Maximum Update Parametrization techniques from natural language models, we provide training guidelines for scaling models to achieve optimal performance. Our results show that within the explored parameter space, clear scaling relationships emerge. These findings suggest that optimal computational resource allocation requires balanced scaling. Specifically, given a tenfold increase in training compute, achieving an optimal seven-fold reduction in mean squared error necessitates an approximately 2.5-fold increase in dataset size and a 3.8-fold increase in model size. This study establishes a foundation for developing spectral foundational models with enhanced domain transfer capabilities.
Authors: Luyao Tang, Yuxuan Yuan, Chaoqi Chen, Zeyu Zhang, Yue Huang, Kun Zhang
Abstract: Although foundation models (FMs) claim to be powerful, their generalization ability significantly decreases when faced with distribution shifts, weak supervision, or malicious attacks in the open world. On the other hand, most domain generalization or adversarial fine-tuning methods are task-related or model-specific, ignoring the universality in practical applications and the transferability between FMs. This paper delves into the problem of generalizing FMs to the out-of-domain data. We propose a novel framework, the Object-Concept-Relation Triad (OCRT), that enables FMs to extract sparse, high-level concepts and intricate relational structures from raw visual inputs. The key idea is to bind objects in visual scenes and a set of object-centric representations through unsupervised decoupling and iterative refinement. To be specific, we project the object-centric representations onto a semantic concept space that the model can readily interpret and estimate their importance to filter out irrelevant elements. Then, a concept-based graph, which has a flexible degree, is constructed to incorporate the set of concepts and their corresponding importance, enabling the extraction of high-order factors from informative concepts and facilitating relational reasoning among these concepts. Extensive experiments demonstrate that OCRT can substantially boost the generalizability and robustness of SAM and CLIP across multiple downstream tasks.
Authors: Chetna Singhal, Yassine Hadjadj-Aoul
Abstract: Efficient network modeling is essential for resource optimization and network planning in next-generation large-scale complex networks. Traditional approaches, such as queuing theory-based modeling and packet-based simulators, can be inefficient due to the assumption made and the computational expense, respectively. To address these challenges, we propose an innovative energy-efficient dynamic orchestration of Graph Neural Networks (GNN) based model training and inference framework for context-aware network modeling and predictions. We have developed a low-complexity solution framework, QAG, that is a Quantum approximation optimization (QAO) algorithm for Adaptive orchestration of GNN-based network modeling. We leverage the tripartite graph model to represent a multi-application system with many compute nodes. Thereafter, we apply the constrained graph-cutting using QAO to find the feasible energy-efficient configurations of the GNN-based model and deploying them on the available compute nodes to meet the network modeling application requirements. The proposed QAG scheme closely matches the optimum and offers atleast a 50% energy saving while meeting the application requirements with 60% lower churn-rate.
Authors: Thomas Sugg, Kyle O'Brien, Lekh Poudel, Alex Dumouchelle, Michelle Jou, Marc Bosch, Deva Ramanan, Srinivasa Narasimhan, Shubham Tulsiani
Abstract: This paper introduces ACC-NVS1, a specialized dataset designed for research on Novel View Synthesis specifically for airborne and ground imagery. Data for ACC-NVS1 was collected in Austin, TX and Pittsburgh, PA in 2023 and 2024. The collection encompasses six diverse real-world scenes captured from both airborne and ground cameras, resulting in a total of 148,000 images. ACC-NVS1 addresses challenges such as varying altitudes and transient objects. This dataset is intended to supplement existing datasets, providing additional resources for comprehensive research, rather than serving as a benchmark.
Authors: Xingwei Liu, Yuexin Chen, Wangli Xu
Abstract: Identification of joint dependence among more than two random vectors plays an important role in many statistical applications, where the data may contain sensitive or confidential information. In this paper, we consider the the d-variable Hilbert-Schmidt independence criterion (dHSIC) in the context of differential privacy. Given the limiting distribution of the empirical estimate of dHSIC is complicated Gaussian chaos, constructing tests in the non-privacy regime is typically based on permutation and bootstrap. To detect joint dependence in privacy, we propose a dHSIC-based testing procedure by employing a differentially private permutation methodology. Our method enjoys privacy guarantee, valid level and pointwise consistency, while the bootstrap counterpart suffers inconsistent power. We further investigate the uniform power of the proposed test in dHSIC metric and $L_2$ metric, indicating that the proposed test attains the minimax optimal power across different privacy regimes. As a byproduct, our results also contain the pointwise and uniform power of the non-private permutation dHSIC, addressing an unsolved question remained in Pfister et al. (2018).
Authors: Nathan Darjana, Ryo Fujii, Hideo Saito, Hiroki Kajita
Abstract: Egocentric open-surgery videos capture rich, fine-grained details essential for accurately modeling surgical procedures and human behavior in the operating room. A detailed, pixel-level understanding of hands and surgical tools is crucial for interpreting a surgeon's actions and intentions. We introduce EgoSurgery-HTS, a new dataset with pixel-wise annotations and a benchmark suite for segmenting surgical tools, hands, and interacting tools in egocentric open-surgery videos. Specifically, we provide a labeled dataset for (1) tool instance segmentation of 14 distinct surgical tools, (2) hand instance segmentation, and (3) hand-tool segmentation to label hands and the tools they manipulate. Using EgoSurgery-HTS, we conduct extensive evaluations of state-of-the-art segmentation methods and demonstrate significant improvements in the accuracy of hand and hand-tool segmentation in egocentric open-surgery videos compared to existing datasets. The dataset will be released at https://github.com/Fujiry0/EgoSurgery.
Authors: Yunus Can G\"ultekin, P\'eter Scheepers, Yuncheng Yuan, Federico Corradi, Alex Alvarado
Abstract: We investigate the design of two neural network (NN) architectures recently proposed as decoders for forward error correction: the so-called single-label NN (SLNN) and multi-label NN (MLNN) decoders. These decoders have been reported to achieve near-optimal codeword- and bit-wise performance, respectively. Results in the literature show near-optimality for a variety of short codes. In this paper, we analytically prove that certain SLNN and MLNN architectures can, in fact, always realize optimal decoding, regardless of the code. These optimal architectures and their binary weights are shown to be defined by the codebook, i.e., no training or network optimization is required. Our proposed architectures are in fact not NNs, but a different way of implementing the maximum likelihood decoding rule. Optimal performance is numerically demonstrated for Hamming $(7,4)$, Polar $(16,8)$, and BCH $(31,21)$ codes. The results show that our optimal architectures are less complex than the SLNN and MLNN architectures proposed in the literature, which in fact only achieve near-optimal performance. Extension to longer codes is still hindered by the curse of dimensionality. Therefore, even though SLNN and MLNN can perform maximum likelihood decoding, such architectures cannot be used for medium and long codes.
Authors: Yan Jia, Harriet Evans, Zoe Porter, Simon Graham, John McDermid, Tom Lawton, David Snead, Ibrahim Habli
Abstract: In this paper we propose an advanced approach to integrating artificial intelligence (AI) into healthcare: autonomous decision support. This approach allows the AI algorithm to act autonomously for a subset of patient cases whilst serving a supportive role in other subsets of patient cases based on defined delegation criteria. By leveraging the complementary strengths of both humans and AI, it aims to deliver greater overall performance than existing human-AI teaming models. It ensures safe handling of patient cases and potentially reduces clinician review time, whilst being mindful of AI tool limitations. After setting the approach within the context of current human-AI teaming models, we outline the delegation criteria and apply them to a specific AI-based tool used in histopathology. The potential impact of the approach and the regulatory requirements for its successful implementation are then discussed.
Authors: Augusto B. Corr\^ea, Andr\'e G. Pereira, Jendrik Seipp
Abstract: In recent years, large language models (LLMs) have shown remarkable capabilities in various artificial intelligence problems. However, they fail to plan reliably, even when prompted with a detailed definition of the planning task. Attempts to improve their planning capabilities, such as chain-of-thought prompting, fine-tuning, and explicit "reasoning" still yield incorrect plans and usually fail to generalize to larger tasks. In this paper, we show how to use LLMs to generate correct plans, even for out-of-distribution tasks of increasing size. For a given planning domain, we ask an LLM to generate several domain-dependent heuristic functions in the form of Python code, evaluate them on a set of training tasks within a greedy best-first search, and choose the strongest one. The resulting LLM-generated heuristics solve many more unseen test tasks than state-of-the-art domain-independent heuristics for classical planning. They are even competitive with the strongest learning algorithm for domain-dependent planning. These findings are especially remarkable given that our proof-of-concept implementation is based on an unoptimized Python planner and the baselines all build upon highly optimized C++ code. In some domains, the LLM-generated heuristics expand fewer states than the baselines, revealing that they are not only efficiently computable, but sometimes even more informative than the state-of-the-art heuristics. Overall, our results show that sampling a set of planning heuristic function programs can significantly improve the planning capabilities of LLMs.
Authors: Markus J. Buehler
Abstract: We report fundamental insights into how agentic graph reasoning systems spontaneously evolve toward a critical state that sustains continuous semantic discovery. By rigorously analyzing structural (Von Neumann graph entropy) and semantic (embedding) entropy, we identify a subtle yet robust regime in which semantic entropy persistently dominates over structural entropy. This interplay is quantified by a dimensionless Critical Discovery Parameter that stabilizes at a small negative value, indicating a consistent excess of semantic entropy. Empirically, we observe a stable fraction (12%) of "surprising" edges, links between semantically distant concepts, providing evidence of long-range or cross-domain connections that drive continuous innovation. Concomitantly, the system exhibits scale-free and small-world topological features, alongside a negative cross-correlation between structural and semantic measures, reinforcing the analogy to self-organized criticality. These results establish clear parallels with critical phenomena in physical, biological, and cognitive complex systems, revealing an entropy-based principle governing adaptability and continuous innovation. Crucially, semantic richness emerges as the underlying driver of sustained exploration, despite not being explicitly used by the reasoning process. Our findings provide interdisciplinary insights and practical strategies for engineering intelligent systems with intrinsic capacities for long-term discovery and adaptation, and offer insights into how model training strategies can be developed that reinforce critical discovery.
Authors: DeShin Hwa, Tobias Holmes, Klaus Drechsler
Abstract: While CNNs were long considered state of the art for image processing, the introduction of Transformer architectures has challenged this position. While achieving excellent results in image classification and segmentation, Transformers remain inherently reliant on large training datasets and remain computationally expensive. A newly introduced Transformer derivative named KV Transformer shows promising results in synthetic, NLP, and image classification tasks, while reducing complexity and memory usage. This is especially conducive to use cases where local inference is required, such as medical screening applications. We endeavoured to further evaluate the merit of KV Transformers on semantic segmentation tasks, specifically in the domain of medical imaging. By directly comparing traditional and KV variants of the same base architectures, we provide further insight into the practical tradeoffs of reduced model complexity. We observe a notable reduction in parameter count and multiply accumulate operations, while achieving similar performance from most of the KV variant models when directly compared to their QKV implementation.
Authors: Chi-Chih Chang, Chien-Yu Lin, Yash Akhauri, Wei-Cheng Lin, Kai-Chiang Wu, Luis Ceze, Mohamed S. Abdelfattah
Abstract: Large Language Models (LLMs) with long context windows enable powerful applications but come at the cost of high memory consumption to store the Key and Value states (KV-Cache). Recent studies attempted to merge KV-cache from multiple layers into shared representations, yet these approaches either require expensive pretraining or rely on assumptions of high per-token cosine similarity across layers which generally does not hold in practice. We find that the dominant singular vectors are remarkably well-aligned across multiple layers of the KV-Cache. Exploiting this insight, we propose xKV, a simple post-training method that applies Singular Value Decomposition (SVD) on the KV-Cache of grouped layers. xKV consolidates the KV-Cache of multiple layers into a shared low-rank subspace, significantly reducing KV-Cache sizes. Through extensive evaluations on the RULER long-context benchmark with widely-used LLMs (e.g., Llama-3.1 and Qwen2.5), xKV achieves up to 6.8x higher compression rates than state-of-the-art inter-layer technique while improving accuracy by 2.7%. Moreover, xKV is compatible with the emerging Multi-Head Latent Attention (MLA) (e.g., DeepSeek-Coder-V2), yielding a notable 3x compression rates on coding tasks without performance degradation. These results highlight xKV's strong capability and versatility in addressing memory bottlenecks for long-context LLM inference. Our code is publicly available at: https://github.com/abdelfattah-lab/xKV.
Authors: Moussa Kassem Sbeyti, Nadja Klein, Azarm Nowzad, Fikret Sivrikaya, Sahin Albayrak
Abstract: Semi-supervised object detection (SSOD) based on pseudo-labeling significantly reduces dependence on large labeled datasets by effectively leveraging both labeled and unlabeled data. However, real-world applications of SSOD often face critical challenges, including class imbalance, label noise, and labeling errors. We present an in-depth analysis of SSOD under real-world conditions, uncovering causes of suboptimal pseudo-labeling and key trade-offs between label quality and quantity. Based on our findings, we propose four building blocks that can be seamlessly integrated into an SSOD framework. Rare Class Collage (RCC): a data augmentation method that enhances the representation of rare classes by creating collages of rare objects. Rare Class Focus (RCF): a stratified batch sampling strategy that ensures a more balanced representation of all classes during training. Ground Truth Label Correction (GLC): a label refinement method that identifies and corrects false, missing, and noisy ground truth labels by leveraging the consistency of teacher model predictions. Pseudo-Label Selection (PLS): a selection method for removing low-quality pseudo-labeled images, guided by a novel metric estimating the missing detection rate while accounting for class rarity. We validate our methods through comprehensive experiments on autonomous driving datasets, resulting in up to 6% increase in SSOD performance. Overall, our investigation and novel, data-centric, and broadly applicable building blocks enable robust and effective SSOD in complex, real-world scenarios. Code is available at https://mos-ks.github.io/publications.
Authors: Shenyuan Gao, Siyuan Zhou, Yilun Du, Jun Zhang, Chuang Gan
Abstract: World models aim to learn action-controlled prediction models and have proven essential for the development of intelligent agents. However, most existing world models rely heavily on substantial action-labeled data and costly training, making it challenging to adapt to novel environments with heterogeneous actions through limited interactions. This limitation can hinder their applicability across broader domains. To overcome this challenge, we propose AdaWorld, an innovative world model learning approach that enables efficient adaptation. The key idea is to incorporate action information during the pretraining of world models. This is achieved by extracting latent actions from videos in a self-supervised manner, capturing the most critical transitions between frames. We then develop an autoregressive world model that conditions on these latent actions. This learning paradigm enables highly adaptable world models, facilitating efficient transfer and learning of new actions even with limited interactions and finetuning. Our comprehensive experiments across multiple environments demonstrate that AdaWorld achieves superior performance in both simulation quality and visual planning.
Authors: Aether Team, Haoyi Zhu, Yifan Wang, Jianjun Zhou, Wenzheng Chang, Yang Zhou, Zizun Li, Junyi Chen, Chunhua Shen, Jiangmiao Pang, Tong He
Abstract: The integration of geometric reconstruction and generative modeling remains a critical challenge in developing AI systems capable of human-like spatial reasoning. This paper proposes Aether, a unified framework that enables geometry-aware reasoning in world models by jointly optimizing three core capabilities: (1) 4D dynamic reconstruction, (2) action-conditioned video prediction, and (3) goal-conditioned visual planning. Through task-interleaved feature learning, Aether achieves synergistic knowledge sharing across reconstruction, prediction, and planning objectives. Building upon video generation models, our framework demonstrates unprecedented synthetic-to-real generalization despite never observing real-world data during training. Furthermore, our approach achieves zero-shot generalization in both action following and reconstruction tasks, thanks to its intrinsic geometric modeling. Remarkably, even without real-world data, its reconstruction performance far exceeds that of domain-specific models. Additionally, Aether leverages a geometry-informed action space to seamlessly translate predictions into actions, enabling effective autonomous trajectory planning. We hope our work inspires the community to explore new frontiers in physically-reasonable world modeling and its applications.
Authors: Aram Davtyan, Sepehr Sameni, Llukman Cerkezi, Givi Meishvilli, Adam Bielski, Paolo Favaro
Abstract: Optimization is often cast as a deterministic problem, where the solution is found through some iterative procedure such as gradient descent. However, when training neural networks the loss function changes over (iteration) time due to the randomized selection of a subset of the samples. This randomization turns the optimization problem into a stochastic one. We propose to consider the loss as a noisy observation with respect to some reference optimum. This interpretation of the loss allows us to adopt Kalman filtering as an optimizer, as its recursive formulation is designed to estimate unknown parameters from noisy measurements. Moreover, we show that the Kalman Filter dynamical model for the evolution of the unknown parameters can be used to capture the gradient dynamics of advanced methods such as Momentum and Adam. We call this stochastic optimization method KOALA, which is short for Kalman Optimization Algorithm with Loss Adaptivity. KOALA is an easy to implement, scalable, and efficient method to train neural networks. We provide convergence analysis and show experimentally that it yields parameter estimates that are on par with or better than existing state of the art optimization algorithms across several neural network architectures and machine learning tasks, such as computer vision and language modeling.
Authors: Ali Raza, Shujun Li, Kim-Phuc Tran, Ludovic Koehl
Abstract: Adversarial attacks such as poisoning attacks have attracted the attention of many machine learning researchers. Traditionally, poisoning attacks attempt to inject adversarial training data in order to manipulate the trained model. In federated learning (FL), data poisoning attacks can be generalized to model poisoning attacks, which cannot be detected by simpler methods due to the lack of access to local training data by the detector. State-of-the-art poisoning attack detection methods for FL have various weaknesses, e.g., the number of attackers has to be known or not high enough, working with i.i.d. data only, and high computational complexity. To overcome above weaknesses, we propose a novel framework for detecting poisoning attacks in FL, which employs a reference model based on a public dataset and an auditor model to detect malicious updates. We implemented a detector based on the proposed framework and using a one-class support vector machine (OC-SVM), which reaches the lowest possible computational complexity O(K) where K is the number of clients. We evaluated our detector's performance against state-of-the-art (SOTA) poisoning attacks for two typical applications of FL: electrocardiograph (ECG) classification and human activity recognition (HAR). Our experimental results validated the performance of our detector over other SOTA detection methods.
Authors: Yuchen Wu, Kangjie Zhou
Abstract: Tensor decomposition serves as a powerful primitive in statistics and machine learning, and has numerous applications in problems such as learning latent variable models or mixture of Gaussians. In this paper, we focus on using power iteration to decompose an overcomplete random tensor. Past work studying the properties of tensor power iteration either requires a non-trivial data-independent initialization, or is restricted to the undercomplete regime. Moreover, several papers implicitly suggest that logarithmically many iterations (in terms of the input dimension) are sufficient for the power method to recover one of the tensor components. Here we present a novel analysis of the dynamics of tensor power iteration from random initialization in the overcomplete regime, where the tensor rank is much greater than its dimension. Surprisingly, we show that polynomially many steps are necessary for convergence of tensor power iteration to any of the true component, which refutes the previous conjecture. On the other hand, our numerical experiments suggest that tensor power iteration successfully recovers tensor components for a broad range of parameters in polynomial time. To further complement our empirical evidence, we prove that a popular objective function for tensor decomposition is strictly increasing along the power iteration path. Our proof is based on the Gaussian conditioning technique, which has been applied to analyze the approximate message passing (AMP) algorithm. The major ingredient of our argument is a conditioning lemma that allows us to generalize AMP-type analysis to non-proportional limit and polynomially many iterations of the power method.
Authors: Xian Yu, Lei Ying
Abstract: Risk-sensitive reinforcement learning (RL) has become a popular tool for controlling the risk of uncertain outcomes and ensuring reliable performance in highly stochastic sequential decision-making problems. While Policy Gradient (PG) methods have been developed for risk-sensitive RL, it remains unclear if these methods enjoy the same global convergence guarantees as in the risk-neutral case \citep{mei2020global,agarwal2021theory,cen2022fast,bhandari2024global}. In this paper, we consider a class of dynamic time-consistent risk measures, named Expected Conditional Risk Measures (ECRMs), and derive PG and Natural Policy Gradient (NPG) updates for ECRMs-based RL problems. We provide global optimality {and iteration complexities} of the proposed algorithms under the following four settings: (i) PG with constrained direct parameterization, (ii) PG with softmax parameterization and log barrier regularization, (iii) NPG with softmax parameterization and entropy regularization, and (iv) approximate NPG with inexact policy evaluation. Furthermore, we test a risk-averse REINFORCE algorithm \citep{williams1992simple} and a risk-averse NPG algorithm \citep{kakade2001natural} on a stochastic Cliffwalk environment to demonstrate the efficacy of our methods and the importance of risk control.
Authors: Tristan S. W. Stevens, Hans van Gorp, Faik C. Meral, Junseob Shin, Jason Yu, Jean-Luc Robert, Ruud J. G. van Sloun
Abstract: Solving ill-posed inverse problems requires careful formulation of prior beliefs over the signals of interest and an accurate description of their manifestation into noisy measurements. Handcrafted signal priors based on e.g. sparsity are increasingly replaced by data-driven deep generative models, and several groups have recently shown that state-of-the-art score-based diffusion models yield particularly strong performance and flexibility. In this paper, we show that the powerful paradigm of posterior sampling with diffusion models can be extended to include rich, structured, noise models. To that end, we propose a joint conditional reverse diffusion process with learned scores for the noise and signal-generating distribution. We demonstrate strong performance gains across various inverse problems with structured noise, outperforming competitive baselines that use normalizing flows and adversarial networks. This opens up new opportunities and relevant practical applications of diffusion modeling for inverse problems in the context of non-Gaussian measurement models.
Authors: Xin Liu, Yaran Chen, Haoran Li, Boyu Li, Dongbin Zhao
Abstract: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible. Unsupervised cross-domain Reinforcement Learning (RL) pre-training shows great potential for challenging continuous visual control but poses a big challenge. In this paper, we propose \textbf{C}ross-domain \textbf{R}andom \textbf{P}re-\textbf{T}raining with \textbf{pro}totypes (CRPTpro), a novel, efficient, and effective self-supervised cross-domain RL pre-training framework. CRPTpro decouples data sampling from encoder pre-training, proposing decoupled random collection to easily and quickly generate a qualified cross-domain pre-training dataset. Moreover, a novel prototypical self-supervised algorithm is proposed to pre-train an effective visual encoder that is generic across different domains. Without finetuning, the cross-domain encoder can be implemented for challenging downstream tasks defined in different domains, either seen or unseen. Compared with recent advanced methods, CRPTpro achieves better performance on downstream policy learning without extra training on exploration agents for data collection, greatly reducing the burden of pre-training. We conduct extensive experiments across eight challenging continuous visual-control domains, including balance control, robot locomotion, and manipulation. CRPTpro significantly outperforms the next best Proto-RL(C) on 11/12 cross-domain downstream tasks with only 54.5\% wall-clock pre-training time, exhibiting state-of-the-art pre-training performance with greatly improved pre-training efficiency.
Authors: Chenguang Wang, Zhang-Hua Fu, Pinyan Lu, Tianshu Yu
Abstract: Efficiently training a multi-task neural solver for various combinatorial optimization problems (COPs) has been less studied so far. Naive application of conventional multi-task learning approaches often falls short in delivering a high-quality, unified neural solver. This deficiency primarily stems from the significant computational demands and a lack of adequate consideration for the complexities inherent in COPs. In this paper, we propose a general and efficient training paradigm to deliver a unified combinatorial multi-task neural solver. To this end, we resort to the theoretical loss decomposition for multiple tasks under an encoder-decoder framework, which enables more efficient training via proper bandit task-sampling algorithms through an intra-task influence matrix. By employing theoretically grounded approximations, our method significantly enhances overall performance, regardless of whether it is within constrained training budgets, across equivalent training epochs, or in terms of generalization capabilities, when compared to conventional training schedules. On the real-world datasets of TSPLib and CVRPLib, our method also achieved the best results compared to single task learning and multi-task learning approaches. Additionally, the influence matrix provides empirical evidence supporting common practices in the field of learning to optimize, further substantiating the effectiveness of our approach. Our code is open-sourced and available at https://github.com/LOGO-CUHKSZ/MTL-COP.
Authors: Thibaud Godon, Baptiste Bauvin, Pascal Germain, Jacques Corbeil, Alexandre Drouin
Abstract: Rule-based models, such as decision trees, appeal to practitioners due to their interpretable nature. However, the learning algorithms that produce such models are often vulnerable to spurious associations and thus, they are not guaranteed to extract causally-relevant insights. In this work, we build on ideas from the invariant causal prediction literature to propose Invariant Causal Set Covering Machines, an extension of the classical Set Covering Machine algorithm for conjunctions/disjunctions of binary-valued rules that provably avoids spurious associations. We demonstrate both theoretically and empirically that our method can identify the causal parents of a variable of interest in polynomial time.
Authors: Jiaming Zhang, Lingyu Qiu, Qi Yi, Yige Li, Jitao Sang, Changsheng Xu, Dit-Yan Yeung
Abstract: The vulnerability of Deep Neural Networks (DNNs) to adversarial attacks poses a significant challenge to their deployment in safety-critical applications. While extensive research has addressed various attack scenarios, the no-box attack setting where adversaries have no prior knowledge, including access to training data of the target model, remains relatively underexplored despite its practical relevance. This work presents a systematic investigation into leveraging large-scale Vision-Language Models (VLMs), particularly CLIP, as surrogate models for executing no-box attacks. Our theoretical and empirical analyses reveal a key limitation in the execution of no-box attacks stemming from insufficient discriminative capabilities for direct application of vanilla CLIP as a surrogate model. To address this limitation, we propose MF-CLIP: a novel framework that enhances CLIP's effectiveness as a surrogate model through margin-aware feature space optimization. Comprehensive evaluations across diverse architectures and datasets demonstrate that MF-CLIP substantially advances the state-of-the-art in no-box attacks, surpassing existing baselines by 15.23% on standard models and achieving a 9.52% improvement on adversarially trained models. Our code will be made publicly available to facilitate reproducibility and future research in this direction.
Authors: Anthony M. Polloreno
Abstract: Reservoir computation is a recurrent framework for learning and predicting time series data, that benefits from extremely simple training and interpretability, often as the the dynamics of a physical system. In this paper, we will study the impact of noise on the learning capabilities of analog reservoir computers. Recent work on reservoir computation has shown that the information processing capacity (IPC) is a useful metric for quantifying the degradation of the performance due to noise. We further this analysis and demonstrate that this degradation of the IPC limits the possible features that can be meaningfully constructed in an analog reservoir computing setting. We borrow a result from quantum complexity theory that relates the circuit model of computation to a continuous time model, and demonstrate an exponential reduction in the accessible volume of reservoir configurations. We conclude by relating this degradation in the IPC to the fat-shattering dimension of a family of functions describing the reservoir dynamics, which allows us to express our result in terms of a classification task. We conclude that any physical, analog reservoir computer that is exposed to noise can only be used to perform a polynomial amount of learning, despite the exponentially large latent space, even with an exponential amount of post-processing.
Authors: Dustin Wright, Christian Igel, Gabrielle Samuel, Raghavendra Selvan
Abstract: Artificial intelligence (AI) is currently spearheaded by machine learning (ML) methods such as deep learning which have accelerated progress on many tasks thought to be out of reach of AI. These recent ML methods are often compute hungry, energy intensive, and result in significant green house gas emissions, a known driver of anthropogenic climate change. Additionally, the platforms on which ML systems run are associated with environmental impacts that go beyond the energy consumption driven carbon emissions. The primary solution lionized by both industry and the ML community to improve the environmental sustainability of ML is to increase the compute and energy efficiency with which ML systems operate. In this perspective, we argue that it is time to look beyond efficiency in order to make ML more environmentally sustainable. We present three high-level discrepancies between the many variables that influence the efficiency of ML and the environmental sustainability of ML. Firstly, we discuss how compute efficiency does not imply energy efficiency or carbon efficiency. Second, we present the unexpected effects of efficiency on operational emissions throughout the ML model life cycle. And, finally, we explore the broader environmental impacts that are not accounted by efficiency. These discrepancies show as to why efficiency alone is not enough to remedy the adverse environmental impacts of ML. Instead, we argue for systems thinking as the next step towards holistically improving the environmental sustainability of ML.
Authors: Yilei Wu, Zijian Dong, Chongyao Chen, Wangchunshu Zhou, Juan Helen Zhou
Abstract: In medical image analysis, regression plays a critical role in computer-aided diagnosis. It enables quantitative measurements such as age prediction from structural imaging, cardiac function quantification, and molecular measurement from PET scans. While deep learning has shown promise for these tasks, most approaches focus solely on optimizing regression loss or model architecture, neglecting the quality of learned feature representations which are crucial for robust clinical predictions. Directly applying representation learning techniques designed for classification to regression often results in fragmented representations in the latent space, yielding sub-optimal performance. In this paper, we argue that the potential of contrastive learning for medical image regression has been overshadowed due to the neglect of two crucial aspects: ordinality-awareness and hardness. To address these challenges, we propose Supervised Contrastive Learning for Medical Imaging Regression with Mixup (SupReMix). It takes anchor-inclusive mixtures (mixup of the anchor and a distinct negative sample) as hard negative pairs and anchor-exclusive mixtures (mixup of two distinct negative samples) as hard positive pairs at the embedding level. This strategy formulates harder contrastive pairs by integrating richer ordinal information. Through theoretical analysis and extensive experiments on six datasets spanning MRI, X-ray, ultrasound, and PET modalities, we demonstrate that SupReMix fosters continuous ordered representations, significantly improving regression performance.
Authors: Parth Paritosh, Nikolay Atanasov, Sonia Martinez
Abstract: In this paper, we aim to design and analyze distributed Bayesian estimation algorithms for sensor networks. The challenges we address are to (i) derive a distributed provably-correct algorithm in the functional space of probability distributions over continuous variables, and (ii) leverage these results to obtain new distributed estimators restricted to subsets of variables observed by individual agents. This relates to applications such as cooperative localization and federated learning, where the data collected at any agent depends on a subset of all variables of interest. We present Bayesian density estimation algorithms using data from non-linear likelihoods at agents in centralized, distributed, and marginal distributed settings. After setting up a distributed estimation objective, we prove almost-sure convergence to the optimal set of pdfs at each agent. Then, we prove the same for a storage-aware algorithm estimating densities only over relevant variables at each agent. Finally, we present a Gaussian version of these algorithms and implement it in a mapping problem using variational inference to handle non-linear likelihood models associated with LiDAR sensing.
Authors: Xiaorui Jiang, Yu Gao, Hengwei Xu, Qi Zhang, Yong Liao, Pengyuan Zhou
Abstract: Federated Learning (FL) allows several clients to cooperatively train machine learning models without disclosing the raw data. In practical applications, asynchronous FL (AFL) can address the straggler effect compared to synchronous FL. However, Non-IID data and stale models pose significant challenges to AFL, as they can diminish the practicality of the global model and even lead to training failures. In this work, we propose a novel AFL framework called Federated Historical Learning (FedHist), which effectively addresses the challenges posed by both Non-IID data and gradient staleness based on the concept of knowledge rumination. FedHist enhances the stability of local gradients by performing weighted fusion with historical global gradients cached on the server. Relying on hindsight, it assigns aggregation weights to each participant in a multi-dimensional manner during each communication round. To further enhance the efficiency and stability of the training process, we introduce an intelligent $\ell_2$-norm amplification scheme, which dynamically regulates the learning progress based on the $\ell_2$-norms of the submitted gradients. Extensive experiments indicate FedHist outperforms state-of-the-art methods in terms of convergence performance and test accuracy.
Authors: Yuchen Wu, Kangjie Zhou
Abstract: We investigate the power iteration algorithm for the tensor PCA model introduced in Richard and Montanari (2014). Previous work studying the properties of tensor power iteration is either limited to a constant number of iterations, or requires a non-trivial data-independent initialization. In this paper, we move beyond these limitations and analyze the dynamics of randomly initialized tensor power iteration up to polynomially many steps. Our contributions are threefold: First, we establish sharp bounds on the number of iterations required for power method to converge to the planted signal, for a broad range of the signal-to-noise ratios. Second, our analysis reveals that the actual algorithmic threshold for power iteration is smaller than the one conjectured in literature by a polylog(n) factor, where n is the ambient dimension. Finally, we propose a simple and effective stopping criterion for power iteration, which provably outputs a solution that is highly correlated with the true signal. Extensive numerical experiments verify our theoretical results.
Authors: Weichen Zhao, Chenguang Wang, Xinyan Wang, Congying Han, Tiande Guo, Tianshu Yu
Abstract: This paper presents an analytical study of the oversmoothing issue in diffusion-based Graph Neural Networks (GNNs). Generalizing beyond extant approaches grounded in random walk analysis or particle systems, we approach this problem through operator semigroup theory. This theoretical framework allows us to rigorously prove that oversmoothing is intrinsically linked to the ergodicity of the diffusion operator. Relying on semigroup method, we can quantitatively analyze the dynamic of graph diffusion and give a specific mathematical form of the smoothing feature by ergodicity and invariant measure of operator, which improves previous works only show existence of oversmoothing. This finding further poses a general and mild ergodicity-breaking condition, encompassing the various specific solutions previously offered, thereby presenting a more universal and theoretically grounded approach to relieve oversmoothing in diffusion-based GNNs. Additionally, we offer a probabilistic interpretation of our theory, forging a link with prior works and broadening the theoretical horizon. Our experimental results reveal that this ergodicity-breaking term effectively mitigates oversmoothing measured by Dirichlet energy, and simultaneously enhances performance in node classification tasks.
Authors: Zitian Li, Wang Chi Cheung
Abstract: Motivated by the cost heterogeneity in experimentation across different alternatives, we study the Best Arm Identification with Resource Constraints (BAIwRC) problem. The agent aims to identify the best arm under resource constraints, where resources are consumed for each arm pull. We make two novel contributions. We design and analyze the Successive Halving with Resource Rationing algorithm (SH-RR). The SH-RR achieves a near-optimal non-asymptotic rate of convergence in terms of the probability of successively identifying an optimal arm. Interestingly, we identify a difference in convergence rates between the cases of deterministic and stochastic resource consumption.
Authors: Cheng Lu, Jiusun Zeng, Yu Xia, Jinhui Cai, Shihua Luo
Abstract: Shapley value is a widely used tool in explainable artificial intelligence (XAI), as it provides a principled way to attribute contributions of input features to model outputs. However, estimation of Shapley value requires capturing conditional dependencies among all feature combinations, which poses significant challenges in complex data environments. In this article, EmSHAP (Energy-based model for Shapley value estimation), an accurate and efficient Shapley value estimation method, is proposed to estimate the expectation of Shapley contribution function under the arbitrary subset of features given the rest. By utilizing the ability of energy-based model (EBM) to model complex distributions, EmSHAP provides an effective solution for estimating the required conditional probabilities. To further improve estimation accuracy, a GRU (Gated Recurrent Unit)-coupled partition function estimation method is introduced. The GRU network captures long-term dependencies with a lightweight parameterization and maps input features into a latent space to mitigate the influence of feature ordering. Additionally, a dynamic masking mechanism is incorporated to further enhance the robustness and accuracy by progressively increasing the masking rate. Theoretical analysis on the error bound as well as application to four case studies verified the higher accuracy and better scalability of EmSHAP in contrast to competitive methods.
Authors: Qingming Li, Juzheng Miao, Puning Zhao, Li Zhou, H. Vicky Zhao, Shouling Ji, Bowen Zhou, Furui Liu
Abstract: In federated learning, client selection is a critical problem that significantly impacts both model performance and fairness. Prior studies typically treat these two objectives separately, or balance them using simple weighting schemes. However, we observe that commonly used metrics for model performance and fairness often conflict with each other, and a straightforward weighted combination is insufficient to capture their complex interactions. To address this, we first propose two guiding principles that directly tackle the inherent conflict between the two metrics while reinforcing each other. Based on these principles, we formulate the client selection problem as a long-term optimization task, leveraging the Lyapunov function and the submodular nature of the problem to solve it effectively. Experiments show that the proposed method improves both model performance and fairness, guiding the system to converge comparably to full client participation. This improvement can be attributed to the fact that both model performance and fairness benefit from the diversity of the selected clients' data distributions. Our approach adaptively enhances this diversity by selecting clients based on their data distributions, thereby improving both model performance and fairness.
Authors: Jihyeon Seong, Sekwang Oh, Jaesik Choi
Abstract: Trend filtering simplifies complex time series data by applying smoothness to filter out noise while emphasizing proximity to the original data. However, existing trend filtering methods fail to reflect abrupt changes in the trend due to `approximateness,' resulting in constant smoothness. This approximateness uniformly filters out the tail distribution of time series data, characterized by extreme values, including both abrupt changes and noise. In this paper, we propose Trend Point Detection formulated as a Markov Decision Process (MDP), a novel approach to identifying essential points that should be reflected in the trend, departing from approximations. We term these essential points as Dynamic Trend Points (DTPs) and extract trends by interpolating them. To identify DTPs, we utilize Reinforcement Learning (RL) within a discrete action space and a forecasting sum-of-squares loss function as a reward, referred to as the Dynamic Trend Filtering network (DTF-net). DTF-net integrates flexible noise filtering, preserving critical original subsequences while removing noise as required for other subsequences. We demonstrate that DTF-net excels at capturing abrupt changes compared to other trend filtering algorithms and enhances forecasting performance, as abrupt changes are predicted rather than smoothed out.
Authors: Matthew Fortier, Mats L. Richter, Oliver Sonnentag, Chris Pal
Abstract: Terrestrial carbon fluxes provide vital information about our biosphere's health and its capacity to absorb anthropogenic CO$_2$ emissions. The importance of predicting carbon fluxes has led to the emerging field of data-driven carbon flux modelling (DDCFM), which uses statistical techniques to predict carbon fluxes from biophysical data. However, the field lacks a standardized dataset to promote comparisons between models. To address this gap, we present CarbonSense, the first machine learning-ready dataset for DDCFM. CarbonSense integrates measured carbon fluxes, meteorological predictors, and satellite imagery from 385 locations across the globe, offering comprehensive coverage and facilitating robust model training. Additionally, we provide a baseline model using a current state-of-the-art DDCFM approach and a novel transformer based model. Our experiments illustrate the potential gains that multimodal deep learning techniques can bring to this domain. By providing these resources, we aim to lower the barrier to entry for other deep learning researchers to develop new models and drive new advances in carbon flux modelling.
Authors: Yanyong Huang, Li Yang, Dongjie Wang, Ke Li, Xiuwen Yi, Fengmao Lv, Tianrui Li
Abstract: Semi-supervised multi-label feature selection has recently been developed to solve the curse of dimensionality problem in high-dimensional multi-label data with certain samples missing labels. Although many efforts have been made, most existing methods use a predefined graph approach to capture the sample similarity or the label correlation. In this manner, the presence of noise and outliers within the original feature space can undermine the reliability of the resulting sample similarity graph. It also fails to precisely depict the label correlation due to the existence of unknown labels. Besides, these methods only consider the discriminative power of selected features, while neglecting their redundancy. In this paper, we propose an Adaptive Collaborative Correlation lEarning-based Semi-Supervised Multi-label Feature Selection (Access-MFS) method to address these issues. Specifically, a generalized regression model equipped with an extended uncorrelated constraint is introduced to select discriminative yet irrelevant features and maintain consistency between predicted and ground-truth labels in labeled data, simultaneously. Then, the instance correlation and label correlation are integrated into the proposed regression model to adaptively learn both the sample similarity graph and the label similarity graph, which mutually enhance feature selection performance. Extensive experimental results demonstrate the superiority of the proposed Access-MFS over other state-of-the-art methods.
Authors: Eviatar Bach, Ricardo Baptista, Enoch Luk, Andrew Stuart
Abstract: Filtering - the task of estimating the conditional distribution for states of a dynamical system given partial and noisy observations - is important in many areas of science and engineering, including weather and climate prediction. However, the filtering distribution is generally intractable to obtain for high-dimensional, nonlinear systems. Filters used in practice, such as the ensemble Kalman filter (EnKF), provide biased probabilistic estimates for nonlinear systems and have numerous tuning parameters. Here, we present a framework for learning a parameterized analysis map - the transformation that takes samples from a forecast distribution, and combines with an observation, to update the approximate filtering distribution - using variational inference. In principle this can lead to a better approximation of the filtering distribution, and hence smaller bias. We show that this methodology can be used to learn the gain matrix, in an affine analysis map, for filtering linear and nonlinear dynamical systems; we also study the learning of inflation and localization parameters for an EnKF. The framework developed here can also be used to learn new filtering algorithms with more general forms for the analysis map.
Authors: Junfan Li, Shizhong Liao
Abstract: Online kernel selection is a fundamental problem of online kernel methods.In this paper,we study online kernel selection with memory constraint in which the memory of kernel selection and online prediction procedures is limited to a fixed budget. An essential question is what is the intrinsic relationship among online learnability, memory constraint, and data complexity? To answer the question,it is necessary to show the trade-offs between regret and memory constraint.Previous work gives a worst-case lower bound depending on the data size,and shows learning is impossible within a small memory constraint.In contrast, we present distinct results by offering data-dependent upper bounds that rely on two data complexities:kernel alignment and the cumulative losses of competitive hypothesis.We propose an algorithmic framework giving data-dependent upper bounds for two types of loss functions.For the hinge loss function,our algorithm achieves an expected upper bound depending on kernel alignment.For smooth loss functions,our algorithm achieves a high-probability upper bound depending on the cumulative losses of competitive hypothesis.We also prove a matching lower bound for smooth loss functions.Our results show that if the two data complexities are sub-linear,then learning is possible within a small memory constraint.Our algorithmic framework depends on a new buffer maintaining framework and a reduction from online kernel selection to prediction with expert advice. Finally,we empirically verify the prediction performance of our algorithms on benchmark datasets.
Authors: Priyank Agrawal, Shipra Agrawal
Abstract: We present an optimistic Q-learning algorithm for regret minimization in average reward reinforcement learning under an additional assumption on the underlying MDP that for all policies, the time to visit some frequent state $s_0$ is finite and upper bounded by $H$, either in expectation or with constant probability. Our setting strictly generalizes the episodic setting and is significantly less restrictive than the assumption of bounded hitting time \textit{for all states} made by most previous literature on model-free algorithms in average reward settings. We demonstrate a regret bound of $\tilde{O}(H^5 S\sqrt{AT})$, where $S$ and $A$ are the numbers of states and actions, and $T$ is the horizon. A key technical novelty of our work is the introduction of an $\overline{L}$ operator defined as $\overline{L} v = \frac{1}{H} \sum_{h=1}^H L^h v$ where $L$ denotes the Bellman operator. Under the given assumption, we show that the $\overline{L}$ operator has a strict contraction (in span) even in the average-reward setting where the discount factor is $1$. Our algorithm design uses ideas from episodic Q-learning to estimate and apply this operator iteratively. Thus, we provide a unified view of regret minimization in episodic and non-episodic settings, which may be of independent interest.
Authors: Thomas Souverain, Johnathan Nguyen, Nicolas Meric, Paul \'Egr\'e
Abstract: In this paper, we propose a philosophical and experimental investigation of the problem of AI fairness in classification. We argue that implementing fairness in AI classification involves more work than just operationalizing a fairness metric. It requires establishing the explainability of the classification model chosen and of the principles behind it. Specifically, it involves making the training processes transparent, determining what outcomes the fairness criteria actually produce, and assessing their trade-offs by comparison with closely related models that would lead to a different outcome. To exemplify this methodology, we trained a model and developed a tool for disparity detection and fairness interventions, the package FairDream. While FairDream is set to enforce Demographic Parity, experiments reveal that it fulfills the constraint of Equalized Odds. The algorithm is thus more conservative than the user might expect. To justify this outcome, we first clarify the relation between Demographic Parity and Equalized Odds as fairness criteria. We then explain FairDream's reweighting method and justify the trade-offs reached by FairDream by a benchmark comparison with closely related GridSearch models. We draw conclusions regarding the way in which these explanatory steps can make an AI model trustworthy.
Authors: Junjing Zheng, Xinyu Zhang, Weidong Jiang, Xiangfeng Qiu, Mingjian Ren
Abstract: Recently, introducing Tensor Decomposition (TD) techniques into unsupervised feature selection (UFS) has been an emerging research topic. A tensor structure is beneficial for mining the relations between different modes and helps relieve the computation burden. However, while existing methods exploit TD to preserve the data tensor structure, they do not consider the influence of data orientation and thus have difficulty in handling orientation-specific data such as time series. To solve the above problem, we utilize the orientation-dependent tensor-tensor product from Tensor Singular Value Decomposition based on *M-product (T-SVDM) and extend the one-dimensional Sparse Principal Component Analysis (SPCA) to a tensor form. The proposed sparse tensor PCA model can constrain sparsity at the specified mode and yield sparse tensor principal components, enhancing flexibility and accuracy in learning feature relations. To ensure fast convergence and a flexible description of feature correlation, we develop a convex version specially designed for general UFS tasks and propose an efficient slice-by-slice algorithm that performs dual optimization in the transform domain. Experimental results on real-world datasets demonstrate the effectiveness and remarkable computational efficiency of the proposed method for tensor data of diverse structures over the state-of-the-arts. With a proper combination of data orientation and transform domain, our method is promising for various applications. The codes related to our proposed methods and the experiments are available at https://github.com/zjj20212035/STPCA.git.
Authors: Zhengbo Wang, Jian Liang, Ran He, Zilei Wang, Tieniu Tan
Abstract: Low-rank adaptation, also known as LoRA, has emerged as a prominent method for parameter-efficient fine-tuning of foundation models. Despite its computational efficiency, LoRA still yields inferior performance compared to full fine-tuning. In this paper, we first uncover a fundamental connection between the optimization processes of LoRA and full fine-tuning: using LoRA for optimization is mathematically equivalent to full fine-tuning using a low-rank gradient for parameter updates. And this low-rank gradient can be expressed in terms of the gradients of the two low-rank matrices in LoRA. Leveraging this insight, we introduce LoRA-Pro, a method that enhances LoRA's performance by strategically adjusting the gradients of these low-rank matrices. This adjustment allows the low-rank gradient to more accurately approximate the full fine-tuning gradient, thereby narrowing the performance gap between LoRA and full fine-tuning. Furthermore, we theoretically derive the optimal solutions for adjusting the gradients of the low-rank matrices, applying them during fine-tuning in LoRA-Pro. We conduct extensive experiments across natural language understanding, dialogue generation, mathematical reasoning, code generation, and image classification tasks, demonstrating that LoRA-Pro substantially improves LoRA's performance, effectively narrowing the gap with full fine-tuning. Code is publicly available at https://github.com/mrflogs/LoRA-Pro.
Authors: Lei Guo, Wei Chen, Yuxuan Sun, Bo Ai, Nikolaos Pappas, Tony Quek
Abstract: Diffusion models have been extensively utilized in AI-generated content (AIGC) in recent years, thanks to the superior generation capabilities. Combining with semantic communications, diffusion models are used for tasks such as denoising, data reconstruction, and content generation. However, existing diffusion-based generative models do not consider the stringent bandwidth limitation, which limits its application in wireless communication. This paper introduces a diffusion-driven semantic communication framework with advanced VAE-based compression for bandwidth-constrained generative model. Our designed architecture utilizes the diffusion model, where the signal transmission process through the wireless channel acts as the forward process in diffusion. To reduce bandwidth requirements, we incorporate a downsampling module and a paired upsampling module based on a variational auto-encoder with reparameterization at the receiver to ensure that the recovered features conform to the Gaussian distribution. Furthermore, we derive the loss function for our proposed system and evaluate its performance through comprehensive experiments. Our experimental results demonstrate significant improvements in pixel-level metrics such as peak signal to noise ratio (PSNR) and semantic metrics like learned perceptual image patch similarity (LPIPS). These enhancements are more profound regarding the compression rates and SNR compared to deep joint source-channel coding (DJSCC).
Authors: Daniel Musekamp, Marimuthu Kalimuthu, David Holzm\"uller, Makoto Takamoto, Mathias Niepert
Abstract: Solving partial differential equations (PDEs) is a fundamental problem in science and engineering. While neural PDE solvers can be more efficient than established numerical solvers, they often require large amounts of training data that is costly to obtain. Active learning (AL) could help surrogate models reach the same accuracy with smaller training sets by querying classical solvers with more informative initial conditions and PDE parameters. While AL is more common in other domains, it has yet to be studied extensively for neural PDE solvers. To bridge this gap, we introduce AL4PDE, a modular and extensible active learning benchmark. It provides multiple parametric PDEs and state-of-the-art surrogate models for the solver-in-the-loop setting, enabling the evaluation of existing and the development of new AL methods for neural PDE solving. We use the benchmark to evaluate batch active learning algorithms such as uncertainty- and feature-based methods. We show that AL reduces the average error by up to 71% compared to random sampling and significantly reduces worst-case errors. Moreover, AL generates similar datasets across repeated runs, with consistent distributions over the PDE parameters and initial conditions. The acquired datasets are reusable, providing benefits for surrogate models not involved in the data generation.
Authors: Yannik Schnitzer, Alessandro Abate, David Parker
Abstract: We present a data-driven approach for producing policies that are provably robust across unknown stochastic environments. Existing approaches can learn models of a single environment as an interval Markov decision processes (IMDP) and produce a robust policy with a probably approximately correct (PAC) guarantee on its performance. However these are unable to reason about the impact of environmental parameters underlying the uncertainty. We propose a framework based on parametric Markov decision processes (MDPs) with unknown distributions over parameters. We learn and analyse IMDPs for a set of unknown sample environments induced by parameters. The key challenge is then to produce meaningful performance guarantees that combine the two layers of uncertainty: (1) multiple environments induced by parameters with an unknown distribution; (2) unknown induced environments which are approximated by IMDPs. We present a novel approach based on scenario optimisation that yields a single PAC guarantee quantifying the risk level for which a specified performance level can be assured in unseen environments, plus a means to trade-off risk and performance. We implement and evaluate our framework using multiple robust policy generation methods on a range of benchmarks. We show that our approach produces tight bounds on a policy's performance with high confidence.
Authors: Qiang Liu, Mengyu Chu, Nils Thuerey
Abstract: The loss functions of many learning problems contain multiple additive terms that can disagree and yield conflicting update directions. For Physics-Informed Neural Networks (PINNs), loss terms on initial/boundary conditions and physics equations are particularly interesting as they are well-established as highly difficult tasks. To improve learning the challenging multi-objective task posed by PINNs, we propose the ConFIG method, which provides conflict-free updates by ensuring a positive dot product between the final update and each loss-specific gradient. It also maintains consistent optimization rates for all loss terms and dynamically adjusts gradient magnitudes based on conflict levels. We additionally leverage momentum to accelerate optimizations by alternating the back-propagation of different loss terms. We provide a mathematical proof showing the convergence of the ConFIG method, and it is evaluated across a range of challenging PINN scenarios. ConFIG consistently shows superior performance and runtime compared to baseline methods. We also test the proposed method in a classic multi-task benchmark, where the ConFIG method likewise exhibits a highly promising performance. Source code is available at https://tum-pbs.github.io/ConFIG
Authors: Aly Lidayan, Michael Dennis, Stuart Russell
Abstract: Intrinsic motivation and reward shaping guide reinforcement learning (RL) agents by adding pseudo-rewards, which can lead to useful emergent behaviors. However, they can also encourage counterproductive exploits, e.g., fixation with noisy TV screens. Here we provide a theoretical model which anticipates these behaviors, and provides broad criteria under which adverse effects can be bounded. We characterize all pseudo-rewards as reward shaping in Bayes-Adaptive Markov Decision Processes (BAMDPs), which formulates the problem of learning in MDPs as an MDP over the agent's knowledge. Optimal exploration maximizes BAMDP state value, which we decompose into the value of the information gathered and the prior value of the physical state. Psuedo-rewards guide RL agents by rewarding behavior that increases these value components, while they hinder exploration when they align poorly with the actual value. We extend potential-based shaping theory to prove BAMDP Potential-based shaping Functions (BAMPFs) are immune to reward-hacking (convergence to behaviors maximizing composite rewards to the detriment of real rewards) in meta-RL, and show empirically how a BAMPF helps a meta-RL agent learn optimal RL algorithms for a Bernoulli Bandit domain. We finally prove that BAMPFs with bounded monotone increasing potentials also resist reward-hacking in the regular RL setting. We show that it is straightforward to retrofit or design new pseudo-reward terms in this form, and provide an empirical demonstration in the Mountain Car environment.
Authors: Zheda Mai, Ping Zhang, Cheng-Hao Tu, Hong-You Chen, Li Zhang, Wei-Lun Chao
Abstract: Parameter-efficient fine-tuning (PEFT) has attracted significant attention due to the growth of pre-trained model sizes and the need to fine-tune (FT) them for superior downstream performance. Despite a surge in new PEFT methods, a systematic study to understand their performance and suitable application scenarios is lacking, leaving questions like "when to apply PEFT" and "which method to use" largely unanswered, especially in visual recognition. In this paper, we conduct a unifying empirical study of representative PEFT methods with Vision Transformers. We systematically tune their hyper-parameters to fairly compare their accuracy on downstream tasks. Our study offers a practical user guide and unveils several new insights. First, if tuned carefully, different PEFT methods achieve similar accuracy in the low-shot benchmark VTAB-1K. This includes simple approaches like FT the bias terms that were reported inferior. Second, despite similar accuracy, we find that PEFT methods make different mistakes and high-confidence predictions, likely due to their different inductive biases. Such an inconsistency (or complementarity) opens up the opportunity for ensemble methods, and we make preliminary attempts at this. Third, going beyond the commonly used low-shot tasks, we find that PEFT is also useful in many-shot regimes, achieving comparable or better accuracy than full FT while using significantly fewer parameters. Lastly, we investigate PEFT's ability to preserve a pre-trained model's robustness to distribution shifts (e.g., CLIP). Perhaps not surprisingly, PEFT approaches outperform full FT alone. However, with weight-space ensembles, full FT can better balance target distribution and distribution shift performance, suggesting a future research direction for robust PEFT.
Authors: Jinhui Bai, Lei Shi
Abstract: Inspired by the structure of spherical harmonics, we propose the truncated kernel stochastic gradient descent (T-kernel SGD) algorithm with a least-square loss function for spherical data fitting. T-kernel SGD introduces a novel regularization strategy by implementing stochastic gradient descent through a closed-form solution of the projection of the stochastic gradient in a low-dimensional subspace. In contrast to traditional kernel SGD, the regularization strategy implemented by T-kernel SGD is more effective in balancing bias and variance by dynamically adjusting the hypothesis space during iterations. The most significant advantage of the proposed algorithm is that it can achieve theoretically optimal convergence rates using a constant step size (independent of the sample size) while overcoming the inherent saturation problem of kernel SGD. Additionally, we leverage the structure of spherical polynomials to derive an equivalent T-kernel SGD, significantly reducing storage and computational costs compared to kernel SGD. Typically, T-kernel SGD requires only $\mathcal{O}(n^{1+\frac{d}{d-1}\epsilon})$ computational complexity and $\mathcal{O}(n^{\frac{d}{d-1}\epsilon})$ storage to achieve optimal rates for the d-dimensional sphere, where $0<\epsilon<\frac{1}{2}$ can be arbitrarily small if the optimal fitting or the underlying space possesses sufficient regularity. This regularity is determined by the smoothness parameter of the objective function and the decaying rate of the eigenvalues of the integral operator associated with the kernel function, both of which reflect the difficulty of the estimation problem. Our main results quantitatively characterize how this prior information influences the convergence of T-kernel SGD. The numerical experiments further validate the theoretical findings presented in this paper.
Authors: Alessandro De Palma, Serge Durand, Zakaria Chihani, Fran\c{c}ois Terrier, Caterina Urban
Abstract: Adversarial training is arguably the most popular way to provide empirical robustness against specific adversarial examples. While variants based on multi-step attacks incur significant computational overhead, single-step variants are vulnerable to a failure mode known as catastrophic overfitting, which hinders their practical utility for large perturbations. A parallel line of work, certified training, has focused on producing networks amenable to formal guarantees of robustness against any possible attack. However, the wide gap between the best-performing empirical and certified defenses has severely limited the applicability of the latter. Inspired by recent developments in certified training, which rely on a combination of adversarial attacks with network over-approximations, and by the connections between local linearity and catastrophic overfitting, we present experimental evidence on the practical utility and limitations of using certified training towards empirical robustness. We show that, when tuned for the purpose, a recent certified training algorithm can prevent catastrophic overfitting on single-step attacks, and that it can bridge the gap to multi-step baselines under appropriate experimental settings. Finally, we present a conceptually simple regularizer for network over-approximations that can achieve similar effects while markedly reducing runtime.
Authors: Adyasha Maharana, Jaehong Yoon, Tianlong Chen, Mohit Bansal
Abstract: Visual instruction datasets from various distributors are released at different times and often contain a significant number of semantically redundant text-image pairs, depending on their task compositions (i.e., skills) or reference sources. This redundancy greatly limits the efficient deployment of continually adaptable multimodal large language models, hindering their ability to refine existing skills and acquire new competencies over time. We reframe the problem of lifelong Instruction Tuning (LiIT) via data selection, where the model automatically selects beneficial samples to learn from earlier and new datasets based on the current state of acquired knowledge in the model. We propose Adapt-$\infty$, a new multi-way and adaptive data selection approach that dynamically balances sample efficiency and effectiveness during LiIT. We first construct pseudo-skill clusters by grouping gradient-based sample vectors. Next, we select the best-performing data selector for each skill cluster from a pool of selector experts, including our newly proposed scoring function, Image Grounding score. This data selector samples a subset of the most important samples from each skill cluster for training. To prevent the continuous increase in the size of the dataset pool during LiIT, we introduce a cluster-wise permanent data pruning strategy to remove the most semantically redundant samples from each cluster, keeping computational requirements manageable. We validate the effectiveness and efficiency of Adapt-$\infty$ over a sequence of multimodal instruction tuning datasets with various tasks, including (Knowledge) VQA, multilingual, grounding, reasoning, language-only, and multi-image comprehension. Training with samples selected by Adapt-$\infty$ alleviates catastrophic forgetting, especially for rare tasks, and promotes forward transfer across the continuum using only a fraction of the original data.
Authors: Severi Rissanen, Markus Heinonen, Arno Solin
Abstract: The covariance for clean data given a noisy observation is an important quantity in many training-free guided generation methods for diffusion models. Current methods require heavy test-time computation, altering the standard diffusion training process or denoiser architecture, or making heavy approximations. We propose a new framework that sidesteps these issues by using covariance information that is available for free from training data and the curvature of the generative trajectory, which is linked to the covariance through the second-order Tweedie's formula. We integrate these sources of information using (i) a novel method to transfer covariance estimates across noise levels and (ii) low-rank updates in a given noise level. We validate the method on linear inverse problems, where it outperforms recent baselines, especially with fewer diffusion steps.
Authors: Arham Khan, Todd Nief, Nathaniel Hudson, Mansi Sakarvadia, Daniel Grzenda, Aswathy Ajith, Jordan Pettyjohn, Kyle Chard, Ian Foster
Abstract: We survey the model merging literature through the lens of loss landscape geometry to connect observations from empirical studies on model merging and loss landscape analysis to phenomena that govern neural network training and the emergence of their inner representations. We distill repeated empirical observations from the literature in these fields into descriptions of four major characteristics of loss landscape geometry: mode convexity, determinism, directedness, and connectivity. We argue that insights into the structure of learned representations from model merging have applications to model interpretability and robustness, subsequently we propose promising new research directions at the intersection of these fields.
Authors: Guangming Huang, Yunfei Long, Cunjin Luo, Sheng Liu
Abstract: Supervised contrastive learning has been explored in making use of label information for multi-label classification, but determining positive samples in multi-label scenario remains challenging. Previous studies have examined strategies for identifying positive samples, considering label overlap proportion between anchors and samples. However, they ignore various relations between given anchors and samples, as well as how to dynamically adjust the weights in contrastive loss functions based on different relations, leading to great ambiguity. In this paper, we introduce five distinct relations between multi-label samples and propose a Similarity-Dissimilarity Loss with contrastive learning for multi-label classification. Our loss function re-weights the loss by computing the similarity and dissimilarity between positive samples and a given anchor based on the introduced relations. We mainly conduct experiments for multi-label text classification on MIMIC datasets, then further extend the evaluation on MS-COCO. The Experimental results show that our proposed loss effectively improves the performance on all encoders under supervised contrastive learning paradigm, demonstrating its effectiveness and robustness.
Authors: Mustofa Ahmed, Abdul Muntakim, Nawrin Tabassum, Mohammad Asifur Rahim, Faisal Muhammad Shah
Abstract: Depression detection using deep learning models has been widely explored in previous studies, especially due to the large amounts of data available from social media posts. These posts provide valuable information about individuals' mental health conditions and can be leveraged to train models and identify patterns in the data. However, distributed learning approaches have not been extensively explored in this domain. In this study, we adopt Federated Learning (FL) to facilitate decentralized training on smartphones while protecting user data privacy. We train three neural network architectures--GRU, RNN, and LSTM on Reddit posts to detect signs of depression and evaluate their performance under heterogeneous FL settings. To optimize the training process, we leverage a common tokenizer across all client devices, which reduces the computational load. Additionally, we analyze resource consumption and communication costs on smartphones to assess their impact in a real-world FL environment. Our experimental results demonstrate that the federated models achieve comparable performance to the centralized models. This study highlights the potential of FL for decentralized mental health prediction by providing a secure and efficient model training process on edge devices.
Authors: Aleksandar Armacki, Shuhua Yu, Dragana Bajovic, Dusan Jakovetic, Soummya Kar
Abstract: We study large deviation upper bounds and mean-squared error (MSE) guarantees of a general framework of nonlinear stochastic gradient methods in the online setting, in the presence of heavy-tailed noise. Unlike existing works that rely on the closed form of a nonlinearity (typically clipping), our framework treats the nonlinearity in a black-box manner, allowing us to provide unified guarantees for a broad class of bounded nonlinearities, including many popular ones, like sign, quantization, normalization, as well as component-wise and joint clipping. We provide several strong results for a broad range of step-sizes in the presence of heavy-tailed noise with symmetric probability density function, positive in a neighbourhood of zero and potentially unbounded moments. In particular, for non-convex costs we provide a large deviation upper bound for the minimum norm-squared of gradients, showing an asymptotic tail decay on an exponential scale, at a rate $\sqrt{t} / \log(t)$. We establish the accompanying rate function, showing an explicit dependence on the choice of step-size, nonlinearity, noise and problem parameters. Next, for non-convex costs and the minimum norm-squared of gradients, we derive the optimal MSE rate $\widetilde{\mathcal{O}}(t^{-1/2})$. Moreover, for strongly convex costs and the last iterate, we provide an MSE rate that can be made arbitrarily close to the optimal rate $\mathcal{O}(t^{-1})$, improving on the state-of-the-art results in the presence of heavy-tailed noise. Finally, we establish almost sure convergence of the minimum norm-squared of gradients, providing an explicit rate, which can be made arbitrarily close to $o(t^{-1/4})$.
Authors: Yikun Bai, Abihith Kothapalli, Hengrong Du, Rocio Diaz Martin, Soheil Kolouri
Abstract: The Gromov-Wasserstein (GW) problem, a variant of the classical optimal transport (OT) problem, has attracted growing interest in the machine learning and data science communities due to its ability to quantify similarity between measures in different metric spaces. However, like the classical OT problem, GW imposes an equal mass constraint between measures, which restricts its application in many machine learning tasks. To address this limitation, the partial Gromov-Wasserstein (PGW) problem has been introduced. It relaxes the equal mass constraint, allowing the comparison of general positive Radon measures. Despite this, both GW and PGW face significant computational challenges due to their non-convex nature. To overcome these challenges, we propose the linear partial Gromov-Wasserstein (LPGW) embedding, a linearized embedding technique for the PGW problem. For $K$ different metric measure spaces, the pairwise computation of the PGW distance requires solving the PGW problem ${O}(K^2)$ times. In contrast, the proposed linearization technique reduces this to ${O}(K)$ times. Similar to the linearization technique for the classical OT problem, we prove that LPGW defines a valid metric for metric measure spaces. Finally, we demonstrate the effectiveness of LPGW in practical applications such as shape retrieval and learning with transport-based embeddings, showing that LPGW preserves the advantages of PGW in partial matching while significantly enhancing computational efficiency. The code is available at https://github.com/mint-vu/Linearized_Partial_Gromov_Wasserstein.
URLs: https://github.com/mint-vu/Linearized_Partial_Gromov_Wasserstein.
Authors: Shi-ang Qi, Yakun Yu, Russell Greiner
Abstract: Survival prediction often involves estimating the time-to-event distribution from censored datasets. Previous approaches have focused on enhancing discrimination and marginal calibration. In this paper, we highlight the significance of conditional calibration for real-world applications -- especially its role in individual decision-making. We propose a method based on conformal prediction that uses the model's predicted individual survival probability at that instance's observed time. This method effectively improves the model's marginal and conditional calibration, without compromising discrimination. We provide asymptotic theoretical guarantees for both marginal and conditional calibration and test it extensively across 15 diverse real-world datasets, demonstrating the method's practical effectiveness and versatility in various settings.
Authors: Zeyuan Li, Yangfan He, Lewei He, Jianhui Wang, Tianyu Shi, Bin Lei, Yuchen Li, Qiuwu Chen
Abstract: Recently, large language models (LLMs) have achieved significant progress in automated code generation. Despite their strong instruction-following capabilities, these models frequently struggled to align with user intent in coding scenarios. In particular, they were hampered by datasets that lacked diversity and failed to address specialized tasks or edge cases. Furthermore, challenges in supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF) led to failures in generating precise, human-intent-aligned code. To tackle these challenges and improve the code generation performance for automated programming systems, we propose Feedback-driven Adaptive Long/short-term memory reinforced Coding Optimization (i.e., FALCON). FALCON is structured into two hierarchical levels. From the global level, long-term memory improves code quality by retaining and applying learned knowledge. At the local level, short-term memory allows for the incorporation of immediate feedback from compilers and AI systems. Additionally, we introduce meta-reinforcement learning with feedback rewards to solve the global-local bi-level optimization problem and enhance the model's adaptability across diverse code generation tasks. Extensive experiments demonstrate that our technique achieves state-of-the-art performance, leading other reinforcement learning methods by more than 4.5 percentage points on the MBPP benchmark and 6.1 percentage points on the Humaneval benchmark. The open-sourced code is publicly available at https://github.com/titurte/FALCON.
Authors: Hsun-Yu Kuo, Yin-Hsiang Liao, Yu-Chieh Chao, Wei-Yun Ma, Pu-Jen Cheng
Abstract: Synthetic data augmentation via large language models (LLMs) allows researchers to leverage additional training data, thus enhancing the performance of downstream tasks, especially when real-world data is scarce. However, the generated data can deviate from the real-world data, and this misalignment can bring deficient outcomes while applying the trained model to applications. Therefore, we proposed efficient weighted-loss approaches to align synthetic data with real-world distribution by emphasizing high-quality and diversified data generated by LLMs with using merely a little real-world data. We empirically assessed the effectiveness of our method on multiple text classification tasks, and the results showed leveraging our approaches on a BERT-level model robustly outperformed standard cross-entropy and other data weighting approaches, providing potential solutions to effectively leveraging synthetic data from any suitable data generator for model training.
Authors: Haiyang Wang, Yue Fan, Muhammad Ferjad Naeem, Yongqin Xian, Jan Eric Lenssen, Liwei Wang, Federico Tombari, Bernt Schiele
Abstract: Transformers have become the predominant architecture in foundation models due to their excellent performance across various domains. However, the substantial cost of scaling these models remains a significant concern. This problem arises primarily from their dependence on a fixed number of parameters within linear projections. When architectural modifications (e.g., channel dimensions) are introduced, the entire model typically requires retraining from scratch. As model sizes continue growing, this strategy results in increasingly high computational costs and becomes unsustainable. To overcome this problem, we introduce TokenFormer, a natively scalable architecture that leverages the attention mechanism not only for computations among input tokens but also for interactions between tokens and model parameters, thereby enhancing architectural flexibility. By treating model parameters as tokens, we replace all the linear projections in Transformers with our token-parameter attention layer, where input tokens act as queries and model parameters as keys and values. This reformulation allows for progressive and efficient scaling without necessitating retraining from scratch. Our model scales from 124M to 1.4B parameters by incrementally adding new key-value parameter pairs, achieving performance comparable to Transformers trained from scratch while greatly reducing training costs. Code and models are available at https://github.com/Haiyang-W/TokenFormer.
Authors: Jinlin Lai, Justin Domke, Daniel Sheldon
Abstract: Bayesian reasoning in linear mixed-effects models (LMMs) is challenging and often requires advanced sampling techniques like Markov chain Monte Carlo (MCMC). A common approach is to write the model in a probabilistic programming language and then sample via Hamiltonian Monte Carlo (HMC). However, there are many ways a user can transform a model that make inference more or less efficient. In particular, marginalizing some variables can greatly improve inference but is difficult for users to do manually. We develop an algorithm to easily marginalize random effects in LMMs. A naive approach introduces cubic time operations within an inference algorithm like HMC, but we reduce the running time to linear using fast linear algebra techniques. We show that marginalization is always beneficial when applicable and highlight improvements in various models, especially ones from cognitive sciences.
Authors: Fardin Jalil Piran, Zhiling Chen, Mohsen Imani, Farhad Imani
Abstract: Federated Learning (FL) has become a key method for preserving data privacy in Internet of Things (IoT) environments, as it trains Machine Learning (ML) models locally while transmitting only model updates. Despite this design, FL remains susceptible to threats such as model inversion and membership inference attacks, which can reveal private training data. Differential Privacy (DP) techniques are often introduced to mitigate these risks, but simply injecting DP noise into black-box ML models can compromise accuracy, particularly in dynamic IoT contexts, where continuous, lifelong learning leads to excessive noise accumulation. To address this challenge, we propose Federated HyperDimensional computing with Privacy-preserving (FedHDPrivacy), an eXplainable Artificial Intelligence (XAI) framework that integrates neuro-symbolic computing and DP. Unlike conventional approaches, FedHDPrivacy actively monitors the cumulative noise across learning rounds and adds only the additional noise required to satisfy privacy constraints. In a real-world application for monitoring manufacturing machining processes, FedHDPrivacy maintains high performance while surpassing standard FL frameworks - Federated Averaging (FedAvg), Federated Proximal (FedProx), Federated Normalized Averaging (FedNova), and Federated Optimization (FedOpt) - by up to 37%. Looking ahead, FedHDPrivacy offers a promising avenue for further enhancements, such as incorporating multimodal data fusion.
Authors: Marcin Rabiza
Abstract: Despite significant advancements in XAI, scholars note a persistent lack of solid conceptual foundations and integration with broader discourse on scientific explanation. In response, emerging XAI research draws on explanatory strategies from various sciences and philosophy of science literature to fill these gaps. This paper outlines a mechanistic strategy for explaining the functional organization of deep learning systems, situating recent advancements in AI explainability within a broader philosophical context. According to the mechanistic approach, the explanation of opaque AI systems involves identifying mechanisms that drive decision-making. For deep neural networks, this means discerning functionally relevant components - such as neurons, layers, circuits, or activation patterns - and understanding their roles through decomposition, localization, and recomposition. Proof-of-principle case studies from image recognition and language modeling align these theoretical approaches with the latest research from AI labs like OpenAI and Anthropic. The paper suggests that such a systematic approach to studying model organization can reveal elements that individual explainability techniques might miss, fostering more thoroughly explainable AI.
Authors: Zikang Zhou, Hengjian Zhou, Haibo Hu, Zihao Wen, Jianping Wang, Yung-Hui Li, Yu-Kai Huang
Abstract: Anticipating the multimodality of future events lays the foundation for safe autonomous driving. However, multimodal motion prediction for traffic agents has been clouded by the lack of multimodal ground truth. Existing works predominantly adopt the winner-take-all training strategy to tackle this challenge, yet still suffer from limited trajectory diversity and uncalibrated mode confidence. While some approaches address these limitations by generating excessive trajectory candidates, they necessitate a post-processing stage to identify the most representative modes, a process lacking universal principles and compromising trajectory accuracy. We are thus motivated to introduce ModeSeq, a new multimodal prediction paradigm that models modes as sequences. Unlike the common practice of decoding multiple plausible trajectories in one shot, ModeSeq requires motion decoders to infer the next mode step by step, thereby more explicitly capturing the correlation between modes and significantly enhancing the ability to reason about multimodality. Leveraging the inductive bias of sequential mode prediction, we also propose the Early-Match-Take-All (EMTA) training strategy to diversify the trajectories further. Without relying on dense mode prediction or heuristic post-processing, ModeSeq considerably improves the diversity of multimodal output while attaining satisfactory trajectory accuracy, resulting in balanced performance on motion prediction benchmarks. Moreover, ModeSeq naturally emerges with the capability of mode extrapolation, which supports forecasting more behavior modes when the future is highly uncertain.
Authors: Seokil Ham, Hee-Seon Kim, Sangmin Woo, Changick Kim
Abstract: Despite the growing interest in Mamba architecture as a potential replacement for Transformer architecture, parameter-efficient fine-tuning (PEFT) approaches for Mamba remain largely unexplored. In our study, we introduce two key insights-driven strategies for PEFT in Mamba architecture: (1) While state-space models (SSMs) have been regarded as the cornerstone of Mamba architecture, then expected to play a primary role in transfer learning, our findings reveal that Projectors -- not SSMs -- are the predominant contributors to transfer learning. (2) Based on our observation, we propose a novel PEFT method specialized to Mamba architecture: Projector-targeted Diagonal-centric Linear Transformation (ProDiaL). ProDiaL focuses on optimizing only the pretrained Projectors for new tasks through diagonal-centric linear transformation matrices, without directly fine-tuning the Projector weights. This targeted approach allows efficient task adaptation, utilizing less than 1% of the total parameters, and exhibits strong performance across both vision and language Mamba models, highlighting its versatility and effectiveness.
Authors: Xianliang Li, Jun Luo, Zhiwei Zheng, Hanxiao Wang, Li Luo, Lingkun Wen, Linlong Wu, Sheng Xu
Abstract: Momentum-based optimizers are widely adopted for training neural networks. However, the optimal selection of momentum coefficients remains elusive. This uncertainty impedes a clear understanding of the role of momentum in stochastic gradient methods. In this paper, we present a frequency domain analysis framework that interprets the momentum method as a time-variant filter for gradients, where adjustments to momentum coefficients modify the filter characteristics. Our experiments support this perspective and provide a deeper understanding of the mechanism involved. Moreover, our analysis reveals the following significant findings: high-frequency gradient components are undesired in the late stages of training; preserving the original gradient in the early stages, and gradually amplifying low-frequency gradient components during training both enhance performance. Based on these insights, we propose Frequency Stochastic Gradient Descent with Momentum (FSGDM), a heuristic optimizer that dynamically adjusts the momentum filtering characteristic with an empirically effective dynamic magnitude response. Experimental results demonstrate the superiority of FSGDM over conventional momentum optimizers.
Authors: Luca Colombo, Fabrizio Pittorino, Manuel Roveri
Abstract: Binary Neural Networks (BNNs) significantly reduce computational complexity and memory usage in machine and deep learning by representing weights and activations with just one bit. However, most existing training algorithms for BNNs rely on quantization-aware floating-point Stochastic Gradient Descent (SGD), limiting the full exploitation of binary operations to the inference phase only. In this work, we propose, for the first time, a fully binary and gradient-free training algorithm for multi-layer BNNs, eliminating the need for back-propagated floating-point gradients. Specifically, the proposed algorithm relies on local binary error signals and binary weight updates, employing integer-valued hidden weights that serve as a synaptic metaplasticity mechanism, thereby enhancing its neurobiological plausibility. The fully binary and gradient-free algorithm introduced in this paper enables the training of binary multi-layer perceptrons with binary inputs, weights, and activations, by using exclusively XNOR, Popcount, and increment/decrement operations. Experimental results on multi-class classification benchmarks show test accuracy improvements of up to +35.47% over the only existing fully binary single-layer state-of-the-art solution. Compared to full-precision SGD, our solution improves test accuracy by up to +41.31% under the same total memory demand$\unicode{x2013}$including the model, activations, and input dataset$\unicode{x2013}$while also reducing computational cost by two orders of magnitude in terms of the total number of equivalent Boolean gates. The proposed algorithm is made available to the scientific community as a public repository.
Authors: Daiheng Zhang, Yan Zeng, Xinyu Hong, Jinbo Xu
Abstract: Accurately predicting protein melting temperature changes (Delta Tm) is fundamental for assessing protein stability and guiding protein engineering. Leveraging multi-modal protein representations has shown great promise in capturing the complex relationships among protein sequences, structures, and functions. In this study, we develop models based on powerful protein language models, including ESM-2, ESM-3 and AlphaFold, using various feature extraction methods to enhance prediction accuracy. By utilizing the ESM-3 model, we achieve a new state-of-the-art performance on the s571 test dataset, obtaining a Pearson correlation coefficient (PCC) of 0.50. Furthermore, we conduct a fair evaluation to compare the performance of different protein language models in the Delta Tm prediction task. Our results demonstrate that integrating multi-modal protein representations could advance the prediction of protein melting temperatures.
Authors: Alex Morehead, Jianlin Cheng
Abstract: Powerful generative AI models of protein-ligand structure have recently been proposed, but few of these methods support both flexible protein-ligand docking and affinity estimation. Of those that do, none can directly model multiple binding ligands concurrently or have been rigorously benchmarked on pharmacologically relevant drug targets, hindering their widespread adoption in drug discovery efforts. In this work, we propose FlowDock, the first deep geometric generative model based on conditional flow matching that learns to directly map unbound (apo) structures to their bound (holo) counterparts for an arbitrary number of binding ligands. Furthermore, FlowDock provides predicted structural confidence scores and binding affinity values with each of its generated protein-ligand complex structures, enabling fast virtual screening of new (multi-ligand) drug targets. For the well-known PoseBusters Benchmark dataset, FlowDock outperforms single-sequence AlphaFold 3 with a 51% blind docking success rate using unbound (apo) protein input structures and without any information derived from multiple sequence alignments, and for the challenging new DockGen-E dataset, FlowDock outperforms single-sequence AlphaFold 3 and matches single-sequence Chai-1 for binding pocket generalization. Additionally, in the ligand category of the 16th community-wide Critical Assessment of Techniques for Structure Prediction (CASP16), FlowDock ranked among the top-5 methods for pharmacological binding affinity estimation across 140 protein-ligand complexes, demonstrating the efficacy of its learned representations in virtual screening. Source code, data, and pre-trained models are available at https://github.com/BioinfoMachineLearning/FlowDock.
Authors: Xiaochong Dong, Jun Dan, Yingyun Sun, Yang Liu, Xuemin Zhang, Shengwei Mei
Abstract: Driven by global climate change and the ongoing energy transition, the coupling between power supply capabilities and meteorological factors has become increasingly significant. Over the long term, accurately quantifying the power generation of renewable energy under the influence of climate change is essential for the development of sustainable power systems. However, due to interdisciplinary differences in data requirements, climate data often lacks the necessary hourly resolution to capture the short-term variability and uncertainties of renewable energy resources. To address this limitation, a super-resolution recurrent diffusion model (SRDM) has been developed to enhance the temporal resolution of climate data and model the short-term uncertainty. The SRDM incorporates a pre-trained decoder and a denoising network, that generates long-term, high-resolution climate data through a recurrent coupling mechanism. The high-resolution climate data is then converted into power value using the mechanism model, enabling the simulation of wind and photovoltaic (PV) power generation on future long-term scales. Case studies were conducted in the Ejina region of Inner Mongolia, China, using fifth-generation reanalysis (ERA5) and coupled model intercomparison project (CMIP6) data under two climate pathways: SSP126 and SSP585. The results demonstrate that the SRDM outperforms existing generative models in generating super-resolution climate data. Furthermore, the research highlights the estimation biases introduced when low-resolution climate data is used for power conversion.
Authors: Alex Mallen, Charlie Griffin, Misha Wagner, Alessandro Abate, Buck Shlegeris
Abstract: An AI control protocol is a plan for usefully deploying AI systems that aims to prevent an AI from intentionally causing some unacceptable outcome. This paper investigates how well AI systems can generate and act on their own strategies for subverting control protocols whilst operating statelessly (without shared memory between contexts). To do this, an AI system may need to reliably generate optimal plans in each context, take actions with well-calibrated probabilities, and coordinate plans with other instances of itself without communicating. We develop Subversion Strategy Eval, a suite of eight environments, covering a range of protocols and strategic capabilities, and six sets of affordances that help isolate individual capabilities. We implement the evaluation in Inspect-AI and release it open-source. We evaluate Claude 3.5 models, including helpful-only versions, as well as OpenAI reasoning models. None of the models demonstrate substantial capability in strategizing to subvert control protocols statelessly. However, providing models with additional affordances, such as the ability to share a plan between contexts, can substantially improve performance. We hope our evaluations can act as a leading indicator for when models are capable of subverting control protocols and also relax the worst-case assumption of perfect strategic ability in AI control evaluations.
Authors: Jiahui Li, Tai-Wei Chang, Kun Kuang, Ximing Li, Long Chen, Jun Zhou
Abstract: Noisy labels are both inevitable and problematic in machine learning methods, as they negatively impact models' generalization ability by causing overfitting. In the context of learning with noise, the transition matrix plays a crucial role in the design of statistically consistent algorithms. However, the transition matrix is often considered unidentifiable. One strand of methods typically addresses this problem by assuming that the transition matrix is instance-independent; that is, the probability of mislabeling a particular instance is not influenced by its characteristics or attributes. This assumption is clearly invalid in complex real-world scenarios. To better understand the transition relationship and relax this assumption, we propose to study the data generation process of noisy labels from a causal perspective. We discover that an unobservable latent variable can affect either the instance itself, the label annotation procedure, or both, which complicates the identification of the transition matrix. To address various scenarios, we have unified these observations within a new causal graph. In this graph, the input instance is divided into a noise-resistant component and a noise-sensitive component based on whether they are affected by the latent variable. These two components contribute to identifying the ``causal transition matrix'', which approximates the true transition matrix with theoretical guarantee. In line with this, we have designed a novel training framework that explicitly models this causal relationship and, as a result, achieves a more accurate model for inferring the clean label.
Authors: Parham Rezaei, Farzan Farnia, Cheuk Ting Li
Abstract: The availability of multiple training algorithms and architectures for generative models requires a selection mechanism to form a single model over a group of well-trained generation models. The selection task is commonly addressed by identifying the model that maximizes an evaluation score based on the diversity and quality of the generated data. However, such a best-model identification approach overlooks the possibility that a mixture of available models can outperform each individual model. In this work, we numerically show that a mixture of generative models on benchmark image datasets can indeed achieve a better evaluation score (based on FID and KID scores), compared to the individual models. This observation motivates the development of efficient algorithms for selecting the optimal mixture of the models. To address this, we formulate a quadratic optimization problem to find an optimal mixture model achieving the maximum of kernel-based evaluation scores including kernel inception distance (KID) and R\'enyi kernel entropy (RKE). To identify the optimal mixture of the models using the fewest possible sample queries, we view the selection task as a multi-armed bandit (MAB) problem and propose the Mixture Upper Confidence Bound (Mixture-UCB) algorithm that provably converges to the optimal mixture of the involved models. More broadly, the proposed Mixture-UCB can be extended to optimize every convex quadratic function of the mixture weights in a general MAB setting. We prove a regret bound for the Mixture-UCB algorithm and perform several numerical experiments to show the success of Mixture-UCB in finding the optimal mixture of text and image generative models. The project code is available at https://github.com/Rezaei-Parham/Mixture-UCB.
Authors: Xianlin Zeng, Yufeng Wang, Yuqi Sun, Guodong Guo, Wenrui Ding, Baochang Zhang
Abstract: Graph Neural Networks (GNNs) have recently gained widespread attention as a successful tool for analyzing graph-structured data. However, imperfect graph structure with noisy links lacks enough robustness and may damage graph representations, therefore limiting the GNNs' performance in practical tasks. Moreover, existing generative architectures fail to fit discriminative graph-related tasks. To tackle these issues, we introduce an unsupervised method based on a joint of generative training and discriminative training to learn graph structure and representation, aiming to improve the discriminative performance of generative models. We propose an Energy-based Contrastive Learning (ECL) guided Graph Structure Refinement (GSR) framework, denoted as ECL-GSR. To our knowledge, this is the first work to combine energy-based models with contrastive learning for GSR. Specifically, we leverage ECL to approximate the joint distribution of sample pairs, which increases the similarity between representations of positive pairs while reducing the similarity between negative ones. Refined structure is produced by augmenting and removing edges according to the similarity metrics among node representations. Extensive experiments demonstrate that ECL-GSR outperforms the state-of-the-art on eight benchmark datasets in node classification. ECL-GSR achieves faster training with fewer samples and memories against the leading baseline, highlighting its simplicity and efficiency in downstream tasks.
Authors: Chia-Yi Hsu, Yu-Lin Tsai, Yu Zhe, Yan-Lun Chen, Chih-Hsun Lin, Chia-Mu Yu, Yang Zhang, Chun-Ying Huang, Jun Sakuma
Abstract: Task arithmetic in large-scale pre-trained models enables agile adaptation to diverse downstream tasks without extensive retraining. By leveraging task vectors (TVs), users can perform modular updates through simple arithmetic operations like addition and subtraction. Yet, this flexibility presents new security challenges. In this paper, we investigate how TVs are vulnerable to backdoor attacks, revealing how malicious actors can exploit them to compromise model integrity. By creating composite backdoors that are designed asymmetrically, we introduce BadTV, a backdoor attack specifically crafted to remain effective simultaneously under task learning, forgetting, and analogy operations. Extensive experiments show that BadTV achieves near-perfect attack success rates across diverse scenarios, posing a serious threat to models relying on task arithmetic. We also evaluate current defenses, finding they fail to detect or mitigate BadTV. Our results highlight the urgent need for robust countermeasures to secure TVs in real-world deployments.
Authors: Ari Blondal, Shan Gao, Hamed Hatami, Pooya Hatami
Abstract: Two seminal papers--Alon, Livni, Malliaris, Moran (STOC 2019) and Bun, Livni, and Moran (FOCS 2020)--established the equivalence between online learnability and globally stable PAC learnability in binary classification. However, Chase, Chornomaz, Moran, and Yehudayoff (STOC 2024) recently showed that this equivalence does not hold in the agnostic setting. Specifically, they proved that in the agnostic setting, only finite hypothesis classes are globally stable learnable. Therefore, agnostic global stability is too restrictive to capture interesting hypothesis classes. To address this limitation, Chase et al. introduced two relaxations of agnostic global stability. In this paper, we characterize the classes that are learnable under their proposed relaxed conditions, resolving the two open problems raised in their work. First, we prove that in the setting where the stability parameter can depend on the excess error (the gap between the learner's error and the best achievable error by the hypothesis class), agnostic stability is fully characterized by the Littlestone dimension. Consequently, as in the realizable case, this form of learnability is equivalent to online learnability. As part of the proof of this theorem, we strengthen the celebrated result of Bun et al. by showing that classes with infinite Littlestone dimension are not stably PAC learnable, even if we allow the stability parameter to depend on the excess error. For the second relaxation proposed by Chase et al., we prove that only finite hypothesis classes are globally stable learnable, even if we restrict the agnostic setting to distributions with small population loss.
Authors: Qi Cheems Wang, Zehao Xiao, Yixiu Mao, Yun Qu, Jiayi Shen, Yiqin Lv, Xiangyang Ji
Abstract: Foundation models have revolutionized general-purpose problem-solving, offering rapid task adaptation through pretraining, meta-training, and finetuning. Recent crucial advances in these paradigms reveal the importance of challenging task prioritized sampling to enhance adaptation robustness under distribution shifts. However, ranking task difficulties over iteration as a preliminary step typically requires exhaustive task evaluation, which is practically unaffordable in computation and data-annotation. This study provides a novel perspective to illuminate the possibility of leveraging the dual importance of adaptation robustness and learning efficiency, particularly in scenarios where task evaluation is risky or costly, such as iterative agent-environment interactions for robotic policy evaluation or computationally intensive inference steps for finetuning foundation models. Firstly, we introduce Model Predictive Task Sampling (MPTS), a framework that bridges the task space and adaptation risk landscape, providing a theoretical foundation for robust active task sampling. MPTS employs a generative model to characterize the episodic optimization process and predicts task-specific adaptation risk via posterior inference. The resulting risk learner amortizes the costly evaluation of task adaptation performance and provably approximates task difficulty rankings. MPTS seamlessly integrates into zero-shot, few-shot, and supervised finetuning settings. Empirically, we conduct extensive experiments in pattern recognition using foundation models and sequential decision-making. Our results demonstrate that MPTS significantly enhances adaptation robustness for tail or out-of-distribution (OOD) tasks and improves learning efficiency compared to state-of-the-art (SOTA) methods. The code is available at the project site https://github.com/thu-rllab/MPTS.
Authors: Zhuang Li
Abstract: The rapid growth of Large Language Models (LLMs) raises concerns about distinguishing AI-generated text from human content. Existing watermarking techniques, like \kgw, struggle with low watermark strength and stringent false-positive requirements. Our analysis reveals that current methods rely on coarse estimates of non-watermarked text, limiting watermark detectability. To address this, we propose Bipolar Watermark (\tool), which splits generated text into positive and negative poles, enhancing detection without requiring additional computational resources or knowledge of the prompt. Theoretical analysis and experimental results demonstrate \tool's effectiveness and compatibility with existing optimization techniques, providing a new optimization dimension for watermarking in LLM-generated content.
Authors: Yang Ji, Ying Sun, Yuting Zhang, Zhigaoyuan Wang, Yuanxin Zhuang, Zheng Gong, Dazhong Shen, Chuan Qin, Hengshu Zhu, Hui Xiong
Abstract: Neural networks have achieved remarkable success across various fields. However, the lack of interpretability limits their practical use, particularly in critical decision-making scenarios. Post-hoc interpretability, which provides explanations for pre-trained models, is often at risk of robustness and fidelity. This has inspired a rising interest in self-interpretable neural networks, which inherently reveal the prediction rationale through the model structures. Although there exist surveys on post-hoc interpretability, a comprehensive and systematic survey of self-interpretable neural networks is still missing. To address this gap, we first collect and review existing works on self-interpretable neural networks and provide a structured summary of their methodologies from five key perspectives: attribution-based, function-based, concept-based, prototype-based, and rule-based self-interpretation. We also present concrete, visualized examples of model explanations and discuss their applicability across diverse scenarios, including image, text, graph data, and deep reinforcement learning. Additionally, we summarize existing evaluation metrics for self-interpretability and identify open challenges in this field, offering insights for future research. To support ongoing developments, we present a publicly accessible resource to track advancements in this domain: https://github.com/yangji721/Awesome-Self-Interpretable-Neural-Network.
URLs: https://github.com/yangji721/Awesome-Self-Interpretable-Neural-Network.
Authors: Kemal Kurniawan, Meladel Mistica, Timothy Baldwin, Jey Han Lau
Abstract: Human label variation (HLV) challenges the standard assumption that a labelled instance has a single ground truth, instead embracing the natural variation in human annotation to train and evaluate models. While various training methods and metrics for HLV have been proposed, it is still unclear which methods and metrics perform best in what settings. We propose new evaluation metrics for HLV leveraging fuzzy set theory. Since these new proposed metrics are differentiable, we then in turn experiment with employing these metrics as training objectives. We conduct an extensive study over 6 HLV datasets testing 14 training methods and 6 evaluation metrics. We find that training on either disaggregated annotations or soft labels performs best across metrics, outperforming training using the proposed training objectives with differentiable metrics. We also show that our proposed soft metric is more interpretable and correlates best with human preference.
Authors: Nghiem T. Diep, Huy Nguyen, Chau Nguyen, Minh Le, Duy M. H. Nguyen, Daniel Sonntag, Mathias Niepert, Nhat Ho
Abstract: The LLaMA-Adapter has recently emerged as an efficient fine-tuning technique for LLaMA models, leveraging zero-initialized attention to stabilize training and enhance performance. However, despite its empirical success, the theoretical foundations of zero-initialized attention remain largely unexplored. In this paper, we provide a rigorous theoretical analysis, establishing a connection between zero-initialized attention and mixture-of-expert models. We prove that both linear and non-linear prompts, along with gating functions, can be optimally estimated, with non-linear prompts offering greater flexibility for future applications. Empirically, we validate our findings on the open LLM benchmarks, demonstrating that non-linear prompts outperform linear ones. Notably, even with limited training data, both prompt types consistently surpass vanilla attention, highlighting the robustness and adaptability of zero-initialized attention.
Authors: Quan Wei, Chung-Yiu Yau, Hoi-To Wai, Yang Katie Zhao, Dongyeop Kang, Youngsuk Park, Mingyi Hong
Abstract: Supervised fine-tuning is a standard method for adapting pre-trained large language models (LLMs) to downstream tasks. Quantization has been recently studied as a post-training technique for efficient LLM deployment. To obtain quantized fine-tuned LLMs, conventional pipelines would first fine-tune the pre-trained models, followed by post-training quantization. This often yields suboptimal performance as it fails to leverage the synergy between fine-tuning and quantization. To effectively realize low-bit quantization of weights, activations, and KV caches in LLMs, we propose an algorithm named Rotated Straight-Through-Estimator (RoSTE), which combines quantization-aware supervised fine-tuning (QA-SFT) with an adaptive rotation strategy that identifies an effective rotation configuration to reduce activation outliers. We provide theoretical insights on RoSTE by analyzing its prediction error when applied to an overparameterized least square quantized training problem. Our findings reveal that the prediction error is directly proportional to the quantization error of the converged weights, which can be effectively managed through an optimized rotation configuration. Experiments on Pythia, Qwen and Llama models of different sizes demonstrate the effectiveness of RoSTE. Compared to existing post-SFT quantization baselines, our method consistently achieves superior performances across various tasks and different LLM architectures.
Authors: Minghong Wu, Minghui Liwang, Yuhan Su, Li Li, Seyyedali Hosseinalipour, Xianbin Wang, Huaiyu Dai, Zhenzhen Jiao
Abstract: Federated Learning (FL) offers a pioneering distributed learning paradigm that enables devices/clients to build a shared global model. This global model is obtained through frequent model transmissions between clients and a central server, which may cause high latency, energy consumption, and congestion over backhaul links. To overcome these drawbacks, Hierarchical Federated Learning (HFL) has emerged, which organizes clients into multiple clusters and utilizes edge nodes (e.g., edge servers) for intermediate model aggregations between clients and the central server. Current research on HFL mainly focus on enhancing model accuracy, latency, and energy consumption in scenarios with a stable/fixed set of clients. However, addressing the dynamic availability of clients -- a critical aspect of real-world scenarios -- remains underexplored. This study delves into optimizing client selection and client-to-edge associations in HFL under intermittent client participation so as to minimize overall system costs (i.e., delay and energy), while achieving fast model convergence. We unveil that achieving this goal involves solving a complex NP-hard problem. To tackle this, we propose a stagewise methodology that splits the solution into two stages, referred to as Plan A and Plan B. Plan A focuses on identifying long-term clients with high chance of participation in subsequent model training rounds. Plan B serves as a backup, selecting alternative clients when long-term clients are unavailable during model training rounds. This stagewise methodology offers a fresh perspective on client selection that can enhance both HFL and conventional FL via enabling low-overhead decision-making processes. Through evaluations on MNIST and CIFAR-10 datasets, we show that our methodology outperforms existing benchmarks in terms of model accuracy and system costs.
Authors: Saul Santos, Ant\'onio Farinhas, Daniel C. McNamee, Andr\'e F. T. Martins
Abstract: Recent research has established a connection between modern Hopfield networks (HNs) and transformer attention heads, with guarantees of exponential storage capacity. However, these models still face challenges scaling storage efficiently. Inspired by psychological theories of continuous neural resource allocation in working memory, we propose an approach that compresses large discrete Hopfield memories into smaller, continuous-time memories. Leveraging continuous attention, our new energy function modifies the update rule of HNs, replacing the traditional softmax-based probability mass function with a probability density, over the continuous memory. This formulation aligns with modern perspectives on human executive function, offering a principled link between attractor dynamics in working memory and resource-efficient memory allocation. Our framework maintains competitive performance with HNs while leveraging a compressed memory, reducing computational costs across synthetic and video datasets.
Authors: Uri Sherman, Tomer Koren, Yishay Mansour
Abstract: Modern policy optimization methods roughly follow the policy mirror descent (PMD) algorithmic template, for which there are by now numerous theoretical convergence results. However, most of these either target tabular environments, or can be applied effectively only when the class of policies being optimized over satisfies strong closure conditions, which is typically not the case when working with parametric policy classes in large-scale environments. In this work, we develop a theoretical framework for PMD for general policy classes where we replace the closure conditions with a strictly weaker variational gradient dominance assumption, and obtain upper bounds on the rate of convergence to the best-in-class policy. Our main result leverages a novel notion of smoothness with respect to a local norm induced by the occupancy measure of the current policy, and casts PMD as a particular instance of smooth non-convex optimization in non-Euclidean space.
Authors: Andreas Maurer
Abstract: The paper gives a bound on the generalization error of the Gibbs algorithm, which recovers known data-independent bounds for the high temperature range and extends to the low-temperature range, where generalization depends critically on the data-dependent loss-landscape. It is shown, that with high probability the generalization error of a single hypothesis drawn from the Gibbs posterior decreases with the total prior volume of all hypotheses with similar or smaller empirical error. This gives theoretical support to the belief in the benefit of flat minima. The zero temperature limit is discussed and the bound is extended to a class of similar stochastic algorithms.
Authors: Krishn Vishwas Kher, Lokesh Venkata Siva Maruthi Badisa, Kusampudi Venkata Datta Sri Harsha, Chitneedi Geetha Sowmya, Saksham Mittal, SakethaNath Jagarlapudi
Abstract: Motivated by the burgeoning interest in cross-domain learning, we present a novel generative modeling challenge: generating counterfactual samples in a target domain based on factual observations from a source domain. Our approach operates within an unsupervised paradigm devoid of parallel or joint datasets, relying exclusively on distinct observational samples and causal graphs for each domain. This setting presents challenges that surpass those of conventional counterfactual generation. Central to our methodology is the disambiguation of exogenous causes into effect-intrinsic and domain-intrinsic categories. This differentiation facilitates the integration of domain-specific causal graphs into a unified joint causal graph via shared effect-intrinsic exogenous variables. We propose leveraging Neural Causal models within this joint framework to enable accurate counterfactual generation under standard identifiability assumptions. Furthermore, we introduce a novel loss function that effectively segregates effect-intrinsic from domain-intrinsic variables during model training. Given a factual observation, our framework combines the posterior distribution of effect-intrinsic variables from the source domain with the prior distribution of domain-intrinsic variables from the target domain to synthesize the desired counterfactuals, adhering to Pearl's causal hierarchy. Intriguingly, when domain shifts are restricted to alterations in causal mechanisms without accompanying covariate shifts, our training regimen parallels the resolution of a conditional optimal transport problem. Empirical evaluations on a synthetic dataset show that our framework generates counterfactuals in the target domain that very closely resemble the ground truth.
Authors: Jiaqi Zhao, Ming Wang, Miao Zhang, Yuzhang Shang, Xuebo Liu, Yaowei Wang, Min Zhang, Liqiang Nie
Abstract: Post-training Quantization (PTQ) technique has been extensively adopted for large language models (LLMs) compression owing to its efficiency and low resource requirement. However, current research lacks a in-depth analysis of the superior and applicable scenarios of each PTQ strategy. In addition, existing algorithms focus primarily on performance, overlooking the trade-off among model size, performance, and quantization bitwidth. To mitigate these confusions, we provide a novel benchmark for LLMs PTQ in this paper. Firstly, in order to support our benchmark, we propose a comprehensive taxonomy for existing mainstream methods by scrutinizing their computational strategies (e.g., optimization-based, compensation-based, etc.). Then, we conduct extensive experiments with the baseline within each class, covering models with various sizes (7B-70B), bitwidths, training levels (LLaMA1/2/3/3.1), architectures (Mixtral, DeepSeekMoE and Mamba) and modality (LLaVA1.5 and VILA1.5) on a wide range of evaluation metrics.Through comparative analysis on the results, we summarize the superior of each PTQ strategy and modelsize-bitwidth trade-off considering the performance. For example, our benchmark reveals that compensation-based technique demonstrates outstanding cross-architecture robustness and extremely low-bit PTQ for ultra large models should be reexamined. Finally, we further accordingly claim that a practical combination of compensation and other PTQ strategy can achieve SOTA various robustness. We believe that our benchmark will provide valuable recommendations for the deployment of LLMs and future research on PTQ approaches.We conduct an repository for our benchmark at https://github.com/zjq0455/PTQ_Benchmark.
Authors: Yushi Zhang, Shuai Su, Yong Wang, Yanzhong Yao
Abstract: This paper develops a novel deep learning approach for solving evolutionary equations, which integrates sequential learning strategies with an enhanced hard constraint strategy featuring trainable parameters, addressing the low computational accuracy of standard Physics-Informed Neural Networks (PINNs) in large temporal domains.Sequential learning strategies divide a large temporal domain into multiple subintervals and solve them one by one in a chronological order, which naturally respects the principle of causality and improves the stability of the PINN solution. The improved hard constraint strategy strictly ensures the continuity and smoothness of the PINN solution at time interval nodes, and at the same time passes the information from the previous interval to the next interval, which avoids the incorrect/trivial solution at the position far from the initial time. Furthermore, by investigating the requirements of different types of equations on hard constraints, we design a novel influence function with trainable parameters for hard constraints, which provides theoretical and technical support for the effective implementations of hard constraint strategies, and significantly improves the universality and computational accuracy of our method. In addition, an adaptive time-domain partitioning algorithm is proposed, which plays an important role in the application of the proposed method as well as in the improvement of computational efficiency and accuracy. Numerical experiments verify the performance of the method. The data and code accompanying this paper are available at https://github.com/zhizhi4452/HCS.
Authors: Jie Ren, Zhenwei Dai, Xianfeng Tang, Hui Liu, Jingying Zeng, Zhen Li, Rahul Goutam, Suhang Wang, Yue Xing, Qi He, Hui Liu
Abstract: Unlearning has been proposed to remove copyrighted and privacy-sensitive data from Large Language Models (LLMs). Existing approaches primarily rely on fine-tuning-based methods, which can be categorized into gradient ascent-based (GA-based) and suppression-based methods. However, they often degrade model utility (the ability to respond to normal prompts). In this work, we aim to develop a general framework that enhances the utility of fine-tuning-based unlearning methods. To achieve this goal, we first investigate the common property between GA-based and suppression-based methods. We unveil that GA-based methods unlearn by distinguishing the target data (i.e., the data to be removed) and suppressing related generations, which is essentially the same strategy employed by suppression-based methods. Inspired by this finding, we introduce Gated Representation UNlearning (GRUN) which has two components: a soft gate function for distinguishing target data and a suppression module using Representation Fine-tuning (ReFT) to adjust representations rather than model parameters. Experiments show that GRUN significantly improves the unlearning and utility. Meanwhile, it is general for fine-tuning-based methods, efficient and promising for sequential unlearning.
Authors: Aristotelis Ballas, Christos Diou
Abstract: Domain Generalization (DG) research has gained considerable traction as of late, since the ability to generalize to unseen data distributions is a requirement that eludes even state-of-the-art training algorithms. In this paper we observe that the initial iterations of model training play a key role in domain generalization effectiveness, since the loss landscape may be significantly different across the training and test distributions, contrary to the case of i.i.d. data. Conflicts between gradients of the loss components of each domain lead the optimization procedure to undesirable local minima that do not capture the domain-invariant features of the target classes. We propose alleviating domain conflicts in model optimization, by iteratively annealing the parameters of a model in the early stages of training and searching for points where gradients align between domains. By discovering a set of parameter values where gradients are updated towards the same direction for each data distribution present in the training set, the proposed Gradient-Guided Annealing (GGA) algorithm encourages models to seek out minima that exhibit improved robustness against domain shifts. The efficacy of GGA is evaluated on five widely accepted and challenging image classification domain generalization benchmarks, where its use alone is able to establish highly competitive or even state-of-the-art performance. Moreover, when combined with previously proposed domain-generalization algorithms it is able to consistently improve their effectiveness by significant margins.
Authors: Andy Gray, Alma Rahat, Tom Crick, Stephen Lindsay
Abstract: Comparative Judgement (CJ) provides an alternative assessment approach by evaluating work holistically rather than breaking it into discrete criteria. This method leverages human ability to make nuanced comparisons, yielding more reliable and valid assessments. CJ aligns with real-world evaluations, where overall quality emerges from the interplay of various elements. However, rubrics remain widely used in education, offering structured criteria for grading and detailed feedback. This creates a gap between CJ's holistic ranking and the need for criterion-based performance breakdowns. This paper addresses this gap using a Bayesian approach. We build on Bayesian CJ (BCJ) by Gray et al., which directly models preferences instead of using likelihoods over total scores, allowing for expected ranks with uncertainty estimation. Their entropy-based active learning method selects the most informative pairwise comparisons for assessors. We extend BCJ to handle multiple independent learning outcome (LO) components, defined by a rubric, enabling both holistic and component-wise predictive rankings with uncertainty estimates. Additionally, we propose a method to aggregate entropies and identify the most informative comparison for assessors. Experiments on synthetic and real data demonstrate our method's effectiveness. Finally, we address a key limitation of BCJ, which is the inability to quantify assessor agreement. We show how to derive agreement levels, enhancing transparency in assessment.
Authors: Jacob Beck
Abstract: While internet-scale image and textual data have enabled strong generalization in Vision-Language Models (VLMs), the absence of internet-scale control data has impeded the development of similar generalization in standard reinforcement learning (RL) agents. Although VLMs are fundamentally limited in their ability to solve control tasks due to their lack of action-conditioned training data, their capacity for image understanding allows them to provide valuable feedback in RL tasks by recognizing successful outcomes. A key challenge in Reinforcement Learning from AI Feedback (RLAIF) is determining how best to integrate VLM-derived signals into the learning process. We explore this question in the context of offline RL and introduce a class of methods called sub-trajectory filtered optimization. We identify three key insights. First, trajectory length plays a crucial role in offline RL, as full-trajectory preference learning exacerbates the stitching problem, necessitating the use of sub-trajectories. Second, even in Markovian environments, a non-Markovian reward signal from a sequence of images is required to assess trajectory improvement, as VLMs do not interpret control actions and must rely on visual cues over time. Third, a simple yet effective approach--filtered and weighted behavior cloning--consistently outperforms more complex reinforcement learning from human feedback-based methods. We propose sub-trajectory filtered behavior cloning, a method that leverages VLM feedback on sub-trajectories while incorporating a retrospective filtering mechanism that removes sub-trajectories preceding failures to improve robustness and prevent turbulence. This study is preliminary; we provide initial evidence through evaluations on a toy control domain. Please enjoy our airport puns.
Authors: Jonathan Geuter, Cl\'ement Bonet, Anna Korba, David Alvarez-Melis
Abstract: Deep Equilibrium Models (DEQs) are a class of implicit neural networks that solve for a fixed point of a neural network in their forward pass. Traditionally, DEQs take sequences as inputs, but have since been applied to a variety of data. In this work, we present Distributional Deep Equilibrium Models (DDEQs), extending DEQs to discrete measure inputs, such as sets or point clouds. We provide a theoretically grounded framework for DDEQs. Leveraging Wasserstein gradient flows, we show how the forward pass of the DEQ can be adapted to find fixed points of discrete measures under permutation-invariance, and derive adequate network architectures for DDEQs. In experiments, we show that they can compete with state-of-the-art models in tasks such as point cloud classification and point cloud completion, while being significantly more parameter-efficient.
Authors: Zhiyin Li, Yubo Yang, Tao Yang, Ziyu Guo, Xiaofeng Wu, Bo Hu
Abstract: Federated learning enables distributed model training across clients without raw data exchange, but in wireless implementations, frequent parameter updates cause high communication overhead. Existing research often assumes known channel state information (CSI) or stationary channels, though practical wireless channels are non-stationary due to fading, user mobility, and attacks, leading to unpredictable transmission failures and exacerbating client staleness, which hampers model convergence. To tackle these challenges, we propose an asynchronous federated learning scheduling framework for non-stationary channels that aims to reduce client staleness while enhancing communication efficiency and fairness. Our framework considers two scenarios: extremely non-stationary and piecewise-stationary channels. Age of Information (AoI) quantifies client staleness under these conditions. We conduct convergence analysis to examine the impact of AoI and per-round client participation on learning performance and formulate the scheduling problem as a multi-armed bandit (MAB) problem. We derive theoretical lower bounds on AoI regret and develop scheduling strategies based on GLR-CUCB and M-exp3 algorithms, including upper bounds on AoI regret. To address imbalanced client updates, we propose an adaptive matching strategy that incorporates marginal utility and fairness considerations. Simulation results show that our algorithm achieves sub-linear AoI regret, accelerates convergence, and promotes fairer aggregation.
Authors: Yingxue Xu, Fengtao Zhou, Chenyu Zhao, Yihui Wang, Can Yang, Hao Chen
Abstract: The integration of multimodal data including pathology images and gene profiles is widely applied in precise survival prediction. Despite recent advances in multimodal survival models, collecting complete modalities for multimodal fusion still poses a significant challenge, hindering their application in clinical settings. Current approaches tackling incomplete modalities often fall short, as they typically compensate for only a limited part of the knowledge of missing modalities. To address this issue, we propose a Distilled Prompt Learning framework (DisPro) to utilize the strong robustness of Large Language Models (LLMs) to missing modalities, which employs two-stage prompting for compensation of comprehensive information for missing modalities. In the first stage, Unimodal Prompting (UniPro) distills the knowledge distribution of each modality, preparing for supplementing modality-specific knowledge of the missing modality in the subsequent stage. In the second stage, Multimodal Prompting (MultiPro) leverages available modalities as prompts for LLMs to infer the missing modality, which provides modality-common information. Simultaneously, the unimodal knowledge acquired in the first stage is injected into multimodal inference to compensate for the modality-specific knowledge of the missing modality. Extensive experiments covering various missing scenarios demonstrated the superiority of the proposed method. The code is available at https://github.com/Innse/DisPro.
Authors: Bassel El Mabsout, Abdelrahman AbdelGawad, Renato Mancuso
Abstract: Practitioners designing reinforcement learning policies face a fundamental challenge: translating intended behavioral objectives into representative reward functions. This challenge stems from behavioral intent requiring simultaneous achievement of multiple competing objectives, typically addressed through labor-intensive linear reward composition that yields brittle results. Consider the ubiquitous robotics scenario where performance maximization directly conflicts with energy conservation. Such competitive dynamics are resistant to simple linear reward combinations. In this paper, we present the concept of objective fulfillment upon which we build Fulfillment Priority Logic (FPL). FPL allows practitioners to define logical formula representing their intentions and priorities within multi-objective reinforcement learning. Our novel Balanced Policy Gradient algorithm leverages FPL specifications to achieve up to 500\% better sample efficiency compared to Soft Actor Critic. Notably, this work constitutes the first implementation of non-linear utility scalarization design, specifically for continuous control problems.
Authors: Yihan Zhou, Eric Price, Trung Nguyen
Abstract: We address the problem of active logistic regression in the realizable setting. It is well known that active learning can require exponentially fewer label queries compared to passive learning, in some cases using $\log \frac{1}{\eps}$ rather than $\poly(1/\eps)$ labels to get error $\eps$ larger than the optimum. We present the first algorithm that is polynomially competitive with the optimal algorithm on every input instance, up to factors polylogarithmic in the error and domain size. In particular, if any algorithm achieves label complexity polylogarithmic in $\eps$, so does ours. Our algorithm is based on efficient sampling and can be extended to learn more general class of functions. We further support our theoretical results with experiments demonstrating performance gains for logistic regression compared to existing active learning algorithms.
Authors: Xiaohong Yang, Minghui Liwang, Liqun Fu, Yuhan Su, Seyyedali Hosseinalipour, Xianbin Wang, Yiguang Hong
Abstract: A key application of HFL lies in smart Internet of Things (IoT) systems, including remote monitoring and battlefield operations, where cellular connectivity is often unavailable. In such scenarios, UAVs can act as mobile aggregators, dynamically providing connectivity to terrestrial IoT devices. Subsequently, this paper investigates an HFL architecture enabled by energy-constrained, dynamically deployed UAVs that are susceptible to communication disruptions. We propose a novel approach to minimize global training costs in such environments by formulating a joint optimization problem that integrates learning configuration, bandwidth allocation, and IoT device-to-UAV association, ensuring timely global aggregation before UAV disconnections and redeployments. The problem explicitly captures the dynamic nature of IoT devices and their intermittent connectivity to UAVs and is shown to be NP-hard. To address its complexity, we decompose the problem into three interrelated subproblems. First, we optimize learning configuration and bandwidth allocation using an augmented Lagrangian function to reduce training costs. Second, we introduce a device fitness score that accounts for data heterogeneity (via Kullback-Leibler divergence), device-to-UAV proximity, and computational resources, leveraging a Twin Delayed Deep Deterministic Policy Gradient (TD3)-based algorithm for adaptive device-to-UAV assignment. Third, we develop a low-complexity two-stage greedy strategy for UAV redeployment and global aggregator selection, ensuring efficient model aggregation despite UAV disconnections.
Authors: Linqi Zhou, Stefano Ermon, Jiaming Song
Abstract: Diffusion models and Flow Matching generate high-quality samples but are slow at inference, and distilling them into few-step models often leads to instability and extensive tuning. To resolve these trade-offs, we propose Inductive Moment Matching (IMM), a new class of generative models for one- or few-step sampling with a single-stage training procedure. Unlike distillation, IMM does not require pre-training initialization and optimization of two networks; and unlike Consistency Models, IMM guarantees distribution-level convergence and remains stable under various hyperparameters and standard model architectures. IMM surpasses diffusion models on ImageNet-256x256 with 1.99 FID using only 8 inference steps and achieves state-of-the-art 2-step FID of 1.98 on CIFAR-10 for a model trained from scratch.
Authors: Zijun Liao, Jinbiao Chen, Debing Wang, Zizhen Zhang, Jiahai Wang
Abstract: Neural Combinatorial Optimization (NCO) has emerged as a promising approach for NP-hard problems. However, prevailing RL-based methods suffer from low sample efficiency due to sparse rewards and underused solutions. We propose Preference Optimization for Combinatorial Optimization (POCO), a training paradigm that leverages solution preferences via objective values. It introduces: (1) an efficient preference pair construction for better explore and exploit solutions, and (2) a novel loss function that adaptively scales gradients via objective differences, removing reliance on reward models or reference policies. Experiments on Job-Shop Scheduling (JSP), Traveling Salesman (TSP), and Flexible Job-Shop Scheduling (FJSP) show POCO outperforms state-of-the-art neural methods, reducing optimality gaps impressively with efficient inference. POCO is architecture-agnostic, enabling seamless integration with existing NCO models, and establishes preference optimization as a principled framework for combinatorial optimization.
Authors: Kyeongkook Seo, Dong-Jun Han, Jaejun Yoo
Abstract: Despite recent advancements in federated learning (FL), the integration of generative models into FL has been limited due to challenges such as high communication costs and unstable training in heterogeneous data environments. To address these issues, we propose PRISM, a FL framework tailored for generative models that ensures (i) stable performance in heterogeneous data distributions and (ii) resource efficiency in terms of communication cost and final model size. The key of our method is to search for an optimal stochastic binary mask for a random network rather than updating the model weights, identifying a sparse subnetwork with high generative performance; i.e., a ``strong lottery ticket''. By communicating binary masks in a stochastic manner, PRISM minimizes communication overhead. This approach, combined with the utilization of maximum mean discrepancy (MMD) loss and a mask-aware dynamic moving average aggregation method (MADA) on the server side, facilitates stable and strong generative capabilities by mitigating local divergence in FL scenarios. Moreover, thanks to its sparsifying characteristic, PRISM yields a lightweight model without extra pruning or quantization, making it ideal for environments such as edge devices. Experiments on MNIST, FMNIST, CelebA, and CIFAR10 demonstrate that PRISM outperforms existing methods, while maintaining privacy with minimal communication costs. PRISM is the first to successfully generate images under challenging non-IID and privacy-preserving FL environments on complex datasets, where previous methods have struggled.
Authors: Zilong Deng, Simon Khan, Shaofeng Zou
Abstract: In this work, we study the sample complexity problem of risk-sensitive Reinforcement Learning (RL) with a generative model, where we aim to maximize the Conditional Value at Risk (CVaR) with risk tolerance level $\tau$ at each step, a criterion we refer to as Iterated CVaR. We first build a connection between Iterated CVaR RL and $(s, a)$-rectangular distributional robust RL with a specific uncertainty set for CVaR. We establish nearly matching upper and lower bounds on the sample complexity of this problem. Specifically, we first prove that a value iteration-based algorithm, ICVaR-VI, achieves an $\epsilon$-optimal policy with at most $\tilde{O} \left(\frac{SA}{(1-\gamma)^4\tau^2\epsilon^2} \right)$ samples, where $\gamma$ is the discount factor, and $S, A$ are the sizes of the state and action spaces. Furthermore, when $\tau \geq \gamma$, the sample complexity improves to $\tilde{O} \left( \frac{SA}{(1-\gamma)^3\epsilon^2} \right)$. We further show a minimax lower bound of $\tilde{O} \left(\frac{(1-\gamma \tau)SA}{(1-\gamma)^4\tau\epsilon^2} \right)$. For a fixed risk level $\tau \in (0,1]$, our upper and lower bounds match, demonstrating the tightness and optimality of our analysis. We also investigate a limiting case with a small risk level $\tau$, called Worst-Path RL, where the objective is to maximize the minimum possible cumulative reward. We develop matching upper and lower bounds of $\tilde{O} \left(\frac{SA}{p_{\min}} \right)$, where $p_{\min}$ denotes the minimum non-zero reaching probability of the transition kernel.
Authors: Yang Ji, Ying Sun, Hengshu Zhu
Abstract: In the era of the knowledge economy, understanding how job skills influence salary is crucial for promoting recruitment with competitive salary systems and aligned salary expectations. Despite efforts on salary prediction based on job positions and talent demographics, there still lacks methods to effectively discern the set-structured skills' intricate composition effect on job salary. While recent advances in neural networks have significantly improved accurate set-based quantitative modeling, their lack of explainability hinders obtaining insights into the skills' composition effects. Indeed, model explanation for set data is challenging due to the combinatorial nature, rich semantics, and unique format. To this end, in this paper, we propose a novel intrinsically explainable set-based neural prototyping approach, namely \textbf{LGDESetNet}, for explainable salary prediction that can reveal disentangled skill sets that impact salary from both local and global perspectives. Specifically, we propose a skill graph-enhanced disentangled discrete subset selection layer to identify multi-faceted influential input subsets with varied semantics. Furthermore, we propose a set-oriented prototype learning method to extract globally influential prototypical sets. The resulting output is transparently derived from the semantic interplay between these input subsets and global prototypes. Extensive experiments on four real-world datasets demonstrate that our method achieves superior performance than state-of-the-art baselines in salary prediction while providing explainable insights into salary-influencing patterns.
Authors: Kevin Wang, Ishaan Javali, Micha{\l} Bortkiewicz, Tomasz Trzci\'nski, Benjamin Eysenbach
Abstract: Scaling up self-supervised learning has driven breakthroughs in language and vision, yet comparable progress has remained elusive in reinforcement learning (RL). In this paper, we study building blocks for self-supervised RL that unlock substantial improvements in scalability, with network depth serving as a critical factor. Whereas most RL papers in recent years have relied on shallow architectures (around 2 - 5 layers), we demonstrate that increasing the depth up to 1024 layers can significantly boost performance. Our experiments are conducted in an unsupervised goal-conditioned setting, where no demonstrations or rewards are provided, so an agent must explore (from scratch) and learn how to maximize the likelihood of reaching commanded goals. Evaluated on simulated locomotion and manipulation tasks, our approach increases performance by $2\times$ - $50\times$. Increasing the model depth not only increases success rates but also qualitatively changes the behaviors learned.
Authors: Tan Le, Van Le
Abstract: We propose the joint graph attention neural network (GAT), clustering with adaptive neighbors (CAN) and probabilistic graphical model for dynamic power flow analysis and fault characteristics. In fact, computational efficiency is the main focus to enhance, whilst we ensure the performance accuracy at the accepted level. Note that Machine Learning (ML) based schemes have a requirement of sufficient labeled data during training, which is not easily satisfied in practical applications. Also, there are unknown data due to new arrived measurements or incompatible smart devices in complex smart grid systems. These problems would be resolved by our proposed GAT based framework, which models the label dependency between the network data and learns object representations such that it could achieve the semi-supervised fault diagnosis. To create the joint label dependency, we develop the graph construction from the raw acquired signals by using CAN. Next, we develop the probabilistic graphical model of Markov random field for graph representation, which supports for the GAT based framework. We then evaluate the proposed framework in the use-case application in smart grid and make a fair comparison to the existing methods.
Authors: Masoud Hashemi, Oluwanifemi Bamgbose, Sathwik Tejaswi Madhusudhan, Jishnu Sethumadhavan Nair, Aman Tiwari, Vikas Yadav
Abstract: Test-time scaling has significantly improved large language model performance, enabling deeper reasoning to solve complex problems. However, this increased reasoning capability also leads to excessive token generation and unnecessary problem-solving attempts. We introduce Don\'t Answer Bench (DNA Bench), a new benchmark designed to evaluate LLMs ability to robustly understand the tricky reasoning triggers and avoiding unnecessary generation. DNA Bench consists of 150 adversarially designed prompts that are easy for humans to understand and respond to, but surprisingly not for many of the recent prominent LLMs. DNA Bench tests models abilities across different capabilities, such as instruction adherence, hallucination avoidance, redundancy filtering, and unanswerable question recognition. We evaluate reasoning LLMs (RLMs), including DeepSeek-R1, OpenAI O3-mini, Claude-3.7-sonnet and compare them against a powerful non-reasoning model, e.g., GPT-4o. Our experiments reveal that RLMs generate up to 70x more tokens than necessary, often failing at tasks that simpler non-reasoning models handle efficiently with higher accuracy. Our findings underscore the need for more effective training and inference strategies in RLMs.
Authors: Jong-Hyun Jeong, Hongki Jo, Qiang Zhou, Tahsin Afroz Hoque Nishat, Lang Wu
Abstract: Wireless sensor networks (WSNs) have become a promising solution for structural health monitoring (SHM), especially in hard-to-reach or remote locations. Battery-powered WSNs offer various advantages over wired systems, however limited battery life has always been one of the biggest obstacles in practical use of the WSNs, regardless of energy harvesting methods. While various methods have been studied for battery health management, existing methods exclusively aim to extend lifetime of individual batteries, lacking a system level view. A consequence of applying such methods is that batteries in a WSN tend to fail at different times, posing significant difficulty on planning and scheduling of battery replacement trip. This study investigate a deep reinforcement learning (DRL) method for active battery degradation management by optimizing duty cycle of WSNs at the system level. This active management strategy effectively reduces earlier failure of battery individuals which enable group replacement without sacrificing WSN performances. A simulated environment based on a real-world WSN setup was developed to train a DRL agent and learn optimal duty cycle strategies. The performance of the strategy was validated in a long-term setup with various network sizes, demonstrating its efficiency and scalability.
Authors: Wei-Chen Wang, Antoine De Comite, Alexandra Voloshina, Monica Daley, Nidhi Seethapathi
Abstract: Modeling movement in real-world tasks is a fundamental scientific goal for motor control, biomechanics, and rehabilitation engineering. However, existing models and their simplifying assumptions such as linear and fixed timescale mappings do not generalize to real-world contexts. Here, we develop a deep learning-based framework for action prediction with architecture-dependent trial embedding, outperforming traditional models across multiple contexts (walking and running, treadmill and overground, varying terrains) and input modalities (multiple body states, gaze). We find that neural network architectures with flexible input history-dependence like GRU and Transformer perform best overall. By quantifying the model's predictions relative to an autoregressive baseline, we identify context- and modality-dependent timescales. There is greater reliance on fast-timescale predictions in complex terrain, gaze predictions precede body state predictions, and full-body state predictions precede center-of-mass-relevant predictions. This deep learning framework for action prediction provides quantifiable insights into the control of complex movements and can be extended to other actions, contexts, and populations.
Authors: Huy Truong, Andr\'es Tello, Alexander Lazovik, Victoria Degeler
Abstract: Privacy restrictions hinder the sharing of real-world Water Distribution Network (WDN) models, limiting the application of emerging data-driven machine learning, which typically requires extensive observations. To address this challenge, we propose the dataset DiTEC-WDN that comprises 36,000 unique scenarios simulated over either short-term (24 hours) or long-term (1 year) periods. We constructed this dataset using an automated pipeline that optimizes crucial parameters (e.g., pressure, flow rate, and demand patterns), facilitates large-scale simulations, and records discrete, synthetic but hydraulically realistic states under standard conditions via rule validation and post-hoc analysis. With a total of 228 million generated graph-based states, DiTEC-WDN can support a variety of machine-learning tasks, including graph-level, node-level, and link-level regression, as well as time-series forecasting. This contribution, released under a public license, encourages open scientific research in the critical water sector, eliminates the risk of exposing sensitive data, and fulfills the need for a large-scale water distribution network benchmark for study comparisons and scenario analysis.
Authors: Eamon Duede
Abstract: This paper aims to clarify the representational status of Deep Learning Models (DLMs). While commonly referred to as 'representations', what this entails is ambiguous due to a conflation of functional and relational conceptions of representation. This paper argues that while DLMs represent their targets in a relational sense, in general, we have no good reason to believe that DLMs encode locally semantically decomposable representations of their targets. That is, the representational capacity these models have is largely global, rather than decomposable into stable, local subrepresentations. This result has immediate implications for explainable AI (XAI) and directs attention toward exploring the global relational nature of deep learning representations and their relationship both to models more generally to understand their potential role in future scientific inquiry.
Authors: Philipp Kuehn, Mike Schmidt, Markus Bayer, Christian Reuter
Abstract: Publicly available information contains valuable information for Cyber Threat Intelligence (CTI). This can be used to prevent attacks that have already taken place on other systems. Ideally, only the initial attack succeeds and all subsequent ones are detected and stopped. But while there are different standards to exchange this information, a lot of it is shared in articles or blog posts in non-standardized ways. Manually scanning through multiple online portals and news pages to discover new threats and extracting them is a time-consuming task. To automize parts of this scanning process, multiple papers propose extractors that use Natural Language Processing (NLP) to extract Indicators of Compromise (IOCs) from documents. However, while this already solves the problem of extracting the information out of documents, the search for these documents is rarely considered. In this paper, a new focused crawler is proposed called ThreatCrawl, which uses Bidirectional Encoder Representations from Transformers (BERT)-based models to classify documents and adapt its crawling path dynamically. While ThreatCrawl has difficulties to classify the specific type of Open Source Intelligence (OSINT) named in texts, e.g., IOC content, it can successfully find relevant documents and modify its path accord ingly. It yields harvest rates of up to 52%, which are, to the best of our knowledge, better than the current state of the art. The results and source code will be made publicly available upon acceptance.
Authors: Xinwei Shen, Peter B\"uhlmann, Armeen Taeb
Abstract: Since distribution shifts are common in real-world applications, there is a pressing need to develop prediction models that are robust against such shifts. Existing frameworks, such as empirical risk minimization or distributionally robust optimization, either lack generalizability for unseen distributions or rely on postulated distance measures. Alternatively, causality offers a data-driven and structural perspective to robust predictions. However, the assumptions necessary for causal inference can be overly stringent, and the robustness offered by such causal models often lacks flexibility. In this paper, we focus on causality-oriented robustness and propose Distributional Robustness via Invariant Gradients (DRIG), a method that exploits general noise interventions in training data for robust predictions against unseen interventions, and naturally interpolates between in-distribution prediction and causality. In a linear setting, we prove that DRIG yields predictions that are robust among a data-dependent class of distribution shifts. Furthermore, we show that our framework includes anchor regression as a special case, and that it yields prediction models that protect against more diverse perturbations. We establish finite-sample results and extend our approach to semi-supervised domain adaptation to further improve prediction performance. Finally, we empirically validate our methods on synthetic simulations and on single-cell and intensive health care datasets.
Authors: R. El Jurdi, G. Varoquaux, O. Colliot
Abstract: Medical segmentation models are evaluated empirically. As such an evaluation is based on a limited set of example images, it is unavoidably noisy. Beyond a mean performance measure, reporting confidence intervals is thus crucial. However, this is rarely done in medical image segmentation. The width of the confidence interval depends on the test set size and on the spread of the performance measure (its standard-deviation across the test set). For classification, many test images are needed to avoid wide confidence intervals. Segmentation, however, has not been studied, and it differs by the amount of information brought by a given test image. In this paper, we study the typical confidence intervals in the context of segmentation in 3D brain magnetic resonance imaging (MRI). We carry experiments on using the standard nnU-net framework, two datasets from the Medical Decathlon challenge that concern brain MRI (hippocampus and brain tumor segmentation) and two performance measures: the Dice Similarity Coefficient and the Hausdorff distance. We show that the parametric confidence intervals are reasonable approximations of the bootstrap estimates for varying test set sizes and spread of the performance metric. Importantly, we show that the test size needed to achieve a given precision is often much lower than for classification tasks. Typically, a 1\% wide confidence interval requires about 100-200 test samples when the spread is low (standard-deviation around 3\%). More difficult segmentation tasks may lead to higher spreads and require over 1000 samples.
Authors: Jonathan W. Siegel
Abstract: We study the following two related problems. The first is to determine to what error an arbitrary zonoid in $\mathbb{R}^{d+1}$ can be approximated in the Hausdorff distance by a sum of $n$ line segments. The second is to determine optimal approximation rates in the uniform norm for shallow ReLU$^k$ neural networks on their variation spaces. The first of these problems has been solved for $d\neq 2,3$, but when $d=2,3$ a logarithmic gap between the best upper and lower bounds remains. We close this gap, which completes the solution in all dimensions. For the second problem, our techniques significantly improve upon existing approximation rates when $k\geq 1$, and enable uniform approximation of both the target function and its derivatives.
Authors: Zhenyu Wang, Peter B\"uhlmann, Zijian Guo
Abstract: Empirical risk minimization often performs poorly when the distribution of the target domain differs from those of source domains. To address such potential distribution shifts, we develop an unsupervised domain adaptation approach that leverages labeled data from multiple source domains and unlabeled data from the target domain. We introduce a distributionally robust model that optimizes an adversarial reward based on the explained variance across a class of target distributions, ensuring generalization to the target domain. We show that the proposed robust model is a weighted average of conditional outcome models from source domains. This formulation allows us to compute the robust model through the aggregation of source models, which can be estimated using various machine learning algorithms of the users' choice, such as random forests, boosting, and neural networks. Additionally, we introduce a bias-correction step to obtain a more accurate aggregation weight, which is effective for various machine learning algorithms. Our framework can be interpreted as a distributionally robust federated learning approach that satisfies privacy constraints while providing insights into the importance of each source for prediction on the target domain. The performance of our method is evaluated on both simulated and real data.
Authors: Ahmad Sajedi, Samir Khaki, Ehsan Amjadian, Lucy Z. Liu, Yuri A. Lawryshyn, Konstantinos N. Plataniotis
Abstract: Researchers have long tried to minimize training costs in deep learning while maintaining strong generalization across diverse datasets. Emerging research on dataset distillation aims to reduce training costs by creating a small synthetic set that contains the information of a larger real dataset and ultimately achieves test accuracy equivalent to a model trained on the whole dataset. Unfortunately, the synthetic data generated by previous methods are not guaranteed to distribute and discriminate as well as the original training data, and they incur significant computational costs. Despite promising results, there still exists a significant performance gap between models trained on condensed synthetic sets and those trained on the whole dataset. In this paper, we address these challenges using efficient Dataset Distillation with Attention Matching (DataDAM), achieving state-of-the-art performance while reducing training costs. Specifically, we learn synthetic images by matching the spatial attention maps of real and synthetic data generated by different layers within a family of randomly initialized neural networks. Our method outperforms the prior methods on several datasets, including CIFAR10/100, TinyImageNet, ImageNet-1K, and subsets of ImageNet-1K across most of the settings, and achieves improvements of up to 6.5% and 4.1% on CIFAR100 and ImageNet-1K, respectively. We also show that our high-quality distilled images have practical benefits for downstream applications, such as continual learning and neural architecture search.
Authors: Brian D. Williamson, Drew King, Ying Huang
Abstract: Estimating a prediction function is a fundamental component of many data analyses. The super learner ensemble, a particular implementation of stacking, has desirable theoretical properties and has been used successfully in many applications. Dimension reduction can be accomplished by using variable screening algorithms (screeners), including the lasso, within the ensemble prior to fitting other prediction algorithms. However, the performance of a super learner using the lasso for dimension reduction has not been fully explored in cases where the lasso is known to perform poorly. We provide empirical results that suggest that a diverse set of candidate screeners should be used to protect against poor performance of any one screener, similar to the guidance for choosing a library of prediction algorithms for the super learner. These results are further illustrated through the analysis of HIV-1 antibody data.
Authors: Mason Ma, Jiajie Wu, Chase Post, Tony Shi, Jingang Yi, Tony Schmitz, Hong Wang
Abstract: This study presents a physics-informed machine learning-based control method for nonlinear dynamic systems with highly noisy measurements. Existing data-driven control methods that use machine learning for system identification cannot effectively cope with highly noisy measurements, resulting in unstable control performance. To address this challenge, the present study extends current physics-informed machine learning capabilities for modeling nonlinear dynamics with control and integrates them into a model predictive control framework. To demonstrate the capability of the proposed method we test and validate with two noisy nonlinear dynamic systems: the chaotic Lorenz 3 system, and turning machine tool. Analysis of the results illustrate that the proposed method outperforms state-of-the-art benchmarks as measured by both modeling accuracy and control performance for nonlinear dynamic systems under high-noise conditions.
Authors: Mario Sanz-Guerrero, Javier Arroyo
Abstract: Peer-to-peer (P2P) lending connects borrowers and lenders through online platforms but suffers from significant information asymmetry, as lenders often lack sufficient data to assess borrowers' creditworthiness. This paper addresses this challenge by leveraging BERT, a Large Language Model (LLM) known for its ability to capture contextual nuances in text, to generate a risk score based on borrowers' loan descriptions using a dataset from the Lending Club platform. We fine-tune BERT to distinguish between defaulted and non-defaulted loans using the loan descriptions provided by the borrowers. The resulting BERT-generated risk score is then integrated as an additional feature into an XGBoost classifier used at the loan granting stage, where decision-makers have limited information available to guide their decisions. This integration enhances predictive performance, with improvements in balanced accuracy and AUC, highlighting the value of textual features in complementing traditional inputs. Moreover, we find that the incorporation of the BERT score alters how classification models utilize traditional input variables, with these changes varying by loan purpose. These findings suggest that BERT discerns meaningful patterns in loan descriptions, encompassing borrower-specific features, specific purposes, and linguistic characteristics. However, the inherent opacity of LLMs and their potential biases underscore the need for transparent frameworks to ensure regulatory compliance and foster trust. Overall, this study demonstrates how LLM-derived insights interact with traditional features in credit risk modeling, opening new avenues to enhance the explainability and fairness of these models.
Authors: Zishi Zhang, Yijie Peng
Abstract: This work seeks to break the sample efficiency bottleneck in parallel large-scale ranking and selection (R&S) problems by leveraging correlation information. We modify the commonly used "divide and conquer" framework in parallel computing by adding a correlation-based clustering step, transforming it into "clustering and conquer". This seemingly simple modification achieves the optimal sample complexity reduction for a widely used class of efficient large-scale R&S procedures. Our approach enjoys two key advantages: 1) it does not require highly accurate correlation estimation or precise clustering, and 2) it allows for seamless integration with various existing R&S procedures, while achieving optimal sample complexity. Theoretically, we develop a novel gradient analysis framework to analyze sample efficiency and guide the design of large-scale R&S procedures. We also introduce a new parallel clustering algorithm tailored for large-scale scenarios. Finally, in large-scale AI applications such as neural architecture search, our methods demonstrate superior performance.
Authors: David R. Burt, Yunyi Shen, Tamara Broderick
Abstract: Spatial prediction tasks are key to weather forecasting, studying air pollution impacts, and other scientific endeavors. Determining how much to trust predictions made by statistical or physical methods is essential for the credibility of scientific conclusions. Unfortunately, classical approaches for validation fail to handle mismatch between locations available for validation and (test) locations where we want to make predictions. This mismatch is often not an instance of covariate shift (as commonly formalized) because the validation and test locations are fixed (e.g., on a grid or at select points) rather than i.i.d. from two distributions. In the present work, we formalize a check on validation methods: that they become arbitrarily accurate as validation data becomes arbitrarily dense. We show that classical and covariate-shift methods can fail this check. We propose a method that builds from existing ideas in the covariate-shift literature, but adapts them to the validation data at hand. We prove that our proposal passes our check. And we demonstrate its advantages empirically on simulated and real data.
Authors: Yuetian Luo, Rina Foygel Barber
Abstract: Algorithm evaluation and comparison are fundamental questions in machine learning and statistics -- how well does an algorithm perform at a given modeling task, and which algorithm performs best? Many methods have been developed to assess algorithm performance, often based around cross-validation type strategies, retraining the algorithm of interest on different subsets of the data and assessing its performance on the held-out data points. Despite the broad use of such procedures, the theoretical properties of these methods are not yet fully understood. In this work, we explore some fundamental limits for answering these questions with limited amounts of data. In particular, we make a distinction between two questions: how good is an algorithm $A$ at the problem of learning from a training set of size $n$, versus, how good is a particular fitted model produced by running $A$ on a particular training data set of size $n$? Our main results prove that, for any test that treats the algorithm $A$ as a ``black box'' (i.e., we can only study the behavior of $A$ empirically), there is a fundamental limit on our ability to carry out inference on the performance of $A$, unless the number of available data points $N$ is many times larger than the sample size $n$ of interest. (On the other hand, evaluating the performance of a particular fitted model is easy as long as a holdout data set is available -- that is, as long as $N-n$ is not too small.) We also ask whether an assumption of algorithmic stability might be sufficient to circumvent this hardness result. Surprisingly, we find that this is not the case: the same hardness result still holds for the problem of evaluating the performance of $A$, aside from a high-stability regime where fitted models are essentially nonrandom. Finally, we also establish similar hardness results for the problem of comparing multiple algorithms.
Authors: Yifan Zhang, Yifan Luo, Yang Yuan, Andrew Chi-Chih Yao
Abstract: We present Autonomous Data Selection (AutoDS), a method that leverages base language models themselves as zero-shot "generative classifiers" to automatically curate high-quality mathematical texts. Unlike prior approaches that require human annotations or training a dedicated data filter, AutoDS relies solely on a model's logits to determine whether a given passage is mathematically informative and educational. By integrating AutoDS into a continual pretraining pipeline, we substantially boost downstream performance on challenging math benchmarks (MATH, GSM8K, and BBH) while using far fewer tokens than previous methods. Empirically, our approach achieves roughly a twofold improvement in pretraining token efficiency over strong baselines, underscoring the potential of self-directed data selection in enhancing mathematical reasoning. We release our curated AutoMathText dataset to facilitate future research in automated domain-specific data curation. The AutoMathText dataset is available at https://huggingface.co/datasets/math-ai/AutoMathText. The code is available at https://github.com/yifanzhang-pro/AutoMathText.
URLs: https://huggingface.co/datasets/math-ai/AutoMathText., https://github.com/yifanzhang-pro/AutoMathText.
Authors: Sheng Liu, Zihan Wang, Yuxiao Chen, Qi Lei
Abstract: Reconstruction attacks and defenses are essential in understanding the data leakage problem in machine learning. However, prior work has centered around empirical observations of gradient inversion attacks, lacks theoretical grounding, and cannot disentangle the usefulness of defending methods from the computational limitation of attacking methods. In this work, we propose to view the problem as an inverse problem, enabling us to theoretically and systematically evaluate the data reconstruction attack. On various defense methods, we derived the algorithmic upper bound and the matching (in feature dimension and architecture dimension) information-theoretical lower bound on the reconstruction error for two-layer neural networks. To complement the theoretical results and investigate the utility-privacy trade-off, we defined a natural evaluation metric of the defense methods with similar utility loss among the strongest attacks. We further propose a strong reconstruction attack that helps update some previous understanding of the strength of defense methods under our proposed evaluation metric.
Authors: Naquee Rizwan, Paramananda Bhaskar, Mithun Das, Swadhin Satyaprakash Majhi, Punyajoy Saha, Animesh Mukherjee
Abstract: There is a rapid increase in the use of multimedia content in current social media platforms. One of the highly popular forms of such multimedia content are memes. While memes have been primarily invented to promote funny and buoyant discussions, malevolent users exploit memes to target individuals or vulnerable communities, making it imperative to identify and address such instances of hateful memes. Thus social media platforms are in dire need for active moderation of such harmful content. While manual moderation is extremely difficult due to the scale of such content, automatic moderation is challenged by the need of good quality annotated data to train hate meme detection algorithms. This makes a perfect pretext for exploring the power of modern day vision language models (VLMs) that have exhibited outstanding performance across various tasks. In this paper we study the effectiveness of VLMs in handling intricate tasks such as hate meme detection in a completely zero-shot setting so that there is no dependency on annotated data for the task. We perform thorough prompt engineering and query state-of-the-art VLMs using various prompt types to detect hateful/harmful memes. We further interpret the misclassification cases using a novel superpixel based occlusion method. Finally we show that these misclassifications can be neatly arranged into a typology of error classes the knowledge of which should enable the design of better safety guardrails in future.
Authors: Han Guo, Ramtin Hosseini, Ruiyi Zhang, Sai Ashish Somayajula, Ranak Roy Chowdhury, Rajesh K. Gupta, Pengtao Xie
Abstract: Masked Autoencoder (MAE) is a notable method for self-supervised pretraining in visual representation learning. It operates by randomly masking image patches and reconstructing these masked patches using the unmasked ones. A key limitation of MAE lies in its disregard for the varying informativeness of different patches, as it uniformly selects patches to mask. To overcome this, some approaches propose masking based on patch informativeness. However, these methods often do not consider the specific requirements of downstream tasks, potentially leading to suboptimal representations for these tasks. In response, we introduce the Multi-level Optimized Mask Autoencoder (MLO-MAE), a novel framework that leverages end-to-end feedback from downstream tasks to learn an optimal masking strategy during pretraining. Our experimental findings highlight MLO-MAE's significant advancements in visual representation learning. Compared to existing methods, it demonstrates remarkable improvements across diverse datasets and tasks, showcasing its adaptability and efficiency.
Authors: Yuang Wang, Siyeop Yoon, Pengfei Jin, Matthew Tivnan, Sifan Song, Zhennong Chen, Rui Hu, Li Zhang, Quanzheng Li, Zhiqiang Chen, Dufan Wu
Abstract: Diffusion-based models have demonstrated remarkable effectiveness in image restoration tasks; however, their iterative denoising process, which starts from Gaussian noise, often leads to slow inference speeds. The Image-to-Image Schr\"odinger Bridge (I$^2$SB) offers a promising alternative by initializing the generative process from corrupted images while leveraging training techniques from score-based diffusion models. In this paper, we introduce the Implicit Image-to-Image Schr\"odinger Bridge (I$^3$SB) to further accelerate the generative process of I$^2$SB. I$^3$SB restructures the generative process into a non-Markovian framework by incorporating the initial corrupted image at each generative step, effectively preserving and utilizing its information. To enable direct use of pretrained I$^2$SB models without additional training, we ensure consistency in marginal distributions. Extensive experiments across many image corruptions, including noise, low resolution, JPEG compression, and sparse sampling, and multiple image modalities, such as natural, human face, and medical images, demonstrate the acceleration benefits of I$^3$SB. Compared to I$^2$SB, I$^3$SB achieves the same perceptual quality with fewer generative steps, while maintaining or improving fidelity to the ground truth.
Authors: Yiming Zhang, Zhuokai Zhao, Zhaorun Chen, Zhili Feng, Zenghui Ding, Yining Sun
Abstract: Self-supervised contrastive learning models, such as CLIP, have set new benchmarks for vision-language models in many downstream tasks. However, their dependency on rigid one-to-one mappings overlooks the complex and often multifaceted relationships between and within texts and images. To this end, we introduce RankCLIP, a novel pre-training method that extends beyond the rigid one-to-one matching framework of CLIP and its variants. By extending the traditional pair-wise loss to list-wise, and leveraging both in-modal and cross-modal ranking consistency, RankCLIP improves the alignment process, enabling it to capture the nuanced many-to-many relationships between and within each modality. Through comprehensive experiments, we demonstrate the effectiveness of RankCLIP in various downstream tasks, notably achieving significant gains in zero-shot classifications over state-of-the-art methods, underscoring the importance of this enhanced learning process.
Authors: Joshua C. Zhao, Saurabh Bagchi, Salman Avestimehr, Kevin S. Chan, Somali Chaterji, Dimitris Dimitriadis, Jiacheng Li, Ninghui Li, Arash Nourian, Holger R. Roth
Abstract: Deep learning has shown incredible potential across a wide array of tasks, and accompanied by this growth has been an insatiable appetite for data. However, a large amount of data needed for enabling deep learning is stored on personal devices, and recent concerns on privacy have further highlighted challenges for accessing such data. As a result, federated learning (FL) has emerged as an important privacy-preserving technology that enables collaborative training of machine learning models without the need to send the raw, potentially sensitive, data to a central server. However, the fundamental premise that sending model updates to a server is privacy-preserving only holds if the updates cannot be "reverse engineered" to infer information about the private training data. It has been shown under a wide variety of settings that this privacy premise does not hold. In this survey paper, we provide a comprehensive literature review of the different privacy attacks and defense methods in FL. We identify the current limitations of these attacks and highlight the settings in which the privacy of an FL client can be broken. We further dissect some of the successful industry applications of FL and draw lessons for future successful adoption. We survey the emerging landscape of privacy regulation for FL and conclude with future directions for taking FL toward the cherished goal of generating accurate models while preserving the privacy of the data from its participants.
Authors: Zebin You, Xinyu Zhang, Hanzhong Guo, Jingdong Wang, Chongxuan Li
Abstract: The ultimate goal of generative models is to perfectly capture the data distribution. For image generation, common metrics of visual quality (e.g., FID) and the perceived truthfulness of generated images seem to suggest that we are nearing this goal. However, through distribution classification tasks, we reveal that, from the perspective of neural network-based classifiers, even advanced diffusion models are still far from this goal. Specifically, classifiers are able to consistently and effortlessly distinguish real images from generated ones across various settings. Moreover, we uncover an intriguing discrepancy: classifiers can easily differentiate between diffusion models with comparable performance (e.g., U-ViT-H vs. DiT-XL), but struggle to distinguish between models within the same family but of different scales (e.g., EDM2-XS vs. EDM2-XXL). Our methodology carries several important implications. First, it naturally serves as a diagnostic tool for diffusion models by analyzing specific features of generated data. Second, it sheds light on the model autophagy disorder and offers insights into the use of generated data: augmenting real data with generated data is more effective than replacing it. Third, classifier guidance can significantly enhance the realism of generated images.
Authors: Deven Sethi, David \v{S}i\v{s}ka, Yufei Zhang
Abstract: Entropy regularization has been widely used in policy optimization algorithms to enhance exploration and the robustness of the optimal control; however it also introduces an additional regularization bias. This work quantifies the impact of entropy regularization on the convergence of policy gradient methods for stochastic exit time control problems. We analyze a continuous-time policy mirror descent dynamics, which updates the policy based on the gradient of an entropy-regularized value function and adjusts the strength of entropy regularization as the algorithm progresses. We prove that with a fixed entropy level, the mirror descent dynamics converges exponentially to the optimal solution of the regularized problem. We further show that when the entropy level decays at suitable polynomial rates, the annealed flow converges to the solution of the unregularized problem at a rate of $\mathcal O(1/S)$ for discrete action spaces and, under suitable conditions, at a rate of $\mathcal O(1/\sqrt{S})$ for general action spaces, with $S$ being the gradient flow running time. The technical challenge lies in analyzing the gradient flow in the infinite-dimensional space of Markov kernels for nonconvex objectives. This paper explains how entropy regularization improves policy optimization, even with the true gradient, from the perspective of convergence rate.
Authors: Seyed Moein Ayyoubzadeh, Kourosh Shahnazari
Abstract: Homograph disambiguation, the task of distinguishing words with identical spellings but different meanings, poses a substantial challenge in natural language processing. In this study, we introduce a novel dataset tailored for Persian homograph disambiguation. Our work encompasses a thorough exploration of various embeddings, evaluated through the cosine similarity method and their efficacy in downstream tasks like classification. Our investigation entails training a diverse array of lightweight machine learning and deep learning models for phonograph disambiguation. We scrutinize the models' performance in terms of Accuracy, Recall, and F1 Score, thereby gaining insights into their respective strengths and limitations. The outcomes of our research underscore three key contributions. First, we present a newly curated Persian dataset, providing a solid foundation for future research in homograph disambiguation. Second, our comparative analysis of embeddings highlights their utility in different contexts, enriching the understanding of their capabilities. Third, by training and evaluating a spectrum of models, we extend valuable guidance for practitioners in selecting suitable strategies for homograph disambiguation tasks. In summary, our study unveils a new dataset, scrutinizes embeddings through diverse perspectives, and benchmarks various models for homograph disambiguation. These findings empower researchers and practitioners to navigate the intricate landscape of homograph-related challenges effectively.
Authors: Baixiang Huang, Yu Luo, Guangyu Wei, Songyan He, Yushuang Shao, Xueying Zeng
Abstract: Coronary artery disease (CAD) is a leading cause of cardiovascular-related mortality, and accurate stenosis detection is crucial for effective clinical decision-making. Coronary angiography remains the gold standard for diagnosing CAD, but manual analysis of angiograms is prone to errors and subjectivity. This study aims to develop a deep learning-based approach for the automatic segmentation of coronary arteries from angiographic images and the quantitative detection of stenosis, thereby improving the accuracy and efficiency of CAD diagnosis. We propose a novel deep learning-based method for the automatic segmentation of coronary arteries in angiographic images, coupled with a dynamic cohort method for stenosis detection. The segmentation model combines the MedSAM and VM-UNet architectures to achieve high-performance results. After segmentation, the vascular centerline is extracted, vessel diameter is computed, and the degree of stenosis is measured with high precision, enabling accurate identification of arterial stenosis. On the mixed dataset (including the ARCADE, DCA1, and GH datasets), the model achieved an average IoU of 0.6308, with sensitivity and specificity of 0.9772 and 0.9903, respectively. On the ARCADE dataset, the average IoU was 0.6303, with sensitivity of 0.9832 and specificity of 0.9933. Additionally, the stenosis detection algorithm achieved a true positive rate (TPR) of 0.5867 and a positive predictive value (PPV) of 0.5911, demonstrating the effectiveness of our model in analyzing coronary angiography images. SAM-VMNet offers a promising tool for the automated segmentation and detection of coronary artery stenosis. The model's high accuracy and robustness provide significant clinical value for the early diagnosis and treatment planning of CAD. The code and examples are available at https://github.com/qimingfan10/SAM-VMNet.
Authors: George I. Austin, Itsik Pe'er, Tal Korem
Abstract: Cross-validation is a common method for estimating the predictive performance of machine learning models. In a data-scarce regime, where one typically wishes to maximize the number of instances used for training the model, an approach called "leave-one-out cross-validation" is often used. In this design, a separate model is built for predicting each data instance after training on all other instances. Since this results in a single test instance available per model trained, predictions are aggregated across the entire dataset to calculate common performance metrics such as the area under the receiver operating characteristic or R2 scores. In this work, we demonstrate that this approach creates a negative correlation between the average label of each training fold and the label of its corresponding test instance, a phenomenon that we term distributional bias. As machine learning models tend to regress to the mean of their training data, this distributional bias tends to negatively impact performance evaluation and hyperparameter optimization. We show that this effect generalizes to leave-P-out cross-validation and persists across a wide range of modeling and evaluation approaches, and that it can lead to a bias against stronger regularization. To address this, we propose a generalizable rebalanced cross-validation approach that corrects for distributional bias for both classification and regression. We demonstrate that our approach improves cross-validation performance evaluation in synthetic simulations, across machine learning benchmarks, and in several published leave-one-out analyses.
Authors: Lane E. Schultz, Yiqi Wang, Ryan Jacobs, Dane Morgan
Abstract: Knowledge of the domain of applicability of a machine learning model is essential to ensuring accurate and reliable model predictions. In this work, we develop a new and general approach of assessing model domain and demonstrate that our approach provides accurate and meaningful domain designation across multiple model types and material property data sets. Our approach assesses the distance between data in feature space using kernel density estimation, where this distance provides an effective tool for domain determination. We show that chemical groups considered unrelated based on chemical knowledge exhibit significant dissimilarities by our measure. We also show that high measures of dissimilarity are associated with poor model performance (i.e., high residual magnitudes) and poor estimates of model uncertainty (i.e., unreliable uncertainty estimation). Automated tools are provided to enable researchers to establish acceptable dissimilarity thresholds to identify whether new predictions of their own machine learning models are in-domain versus out-of-domain.
Authors: Tian Liu, Huixin Zhang, Shubham Parashar, Shu Kong
Abstract: Few-shot recognition (FSR) aims to train a classification model with only a few labeled examples of each concept concerned by a downstream task, where data annotation cost can be prohibitively high. We develop methods to solve FSR by leveraging a pretrained Vision-Language Model (VLM). We particularly explore retrieval-augmented learning (RAL), which retrieves open data, e.g., the VLM's pretraining dataset, to learn models for better serving downstream tasks. RAL has been studied in zero-shot recognition but remains under-explored in FSR. Although applying RAL to FSR may seem straightforward, we observe interesting and novel challenges and opportunities. First, somewhat surprisingly, finetuning a VLM on a large amount of retrieved data underperforms state-of-the-art zero-shot methods. This is due to the imbalanced distribution of retrieved data and its domain gaps with the few-shot examples in the downstream task. Second, more surprisingly, we find that simply finetuning a VLM solely on few-shot examples significantly outperforms previous FSR methods, and finetuning on the mix of retrieved and few-shot data yields even better results. Third, to mitigate the imbalanced distribution and domain gap issues, we propose Stage-Wise retrieval-Augmented fineTuning (SWAT), which involves end-to-end finetuning on mixed data in the first stage and retraining the classifier on the few-shot data in the second stage. Extensive experiments on nine popular benchmarks demonstrate that SWAT significantly outperforms previous methods by >6% accuracy.
Authors: Shihua Sun, Shridatt Sugrim, Angelos Stavrou, Haining Wang
Abstract: Federated Learning (FL) exposes vulnerabilities to targeted poisoning attacks that aim to cause misclassification specifically from the source class to the target class. However, using well-established defense frameworks, the poisoning impact of these attacks can be greatly mitigated. We introduce a generalized pre-training stage approach to Boost Targeted Poisoning Attacks against FL, called BoTPA. Its design rationale is to leverage the model update contributions of all data points, including ones outside of the source and target classes, to construct an Amplifier set, in which we falsify the data labels before the FL training process, as a means to boost attacks. We comprehensively evaluate the effectiveness and compatibility of BoTPA on various targeted poisoning attacks. Under data poisoning attacks, our evaluations reveal that BoTPA can achieve a median Relative Increase in Attack Success Rate (RI-ASR) between 15.3% and 36.9% across all possible source-target class combinations, with varying percentages of malicious clients, compared to its baseline. In the context of model poisoning, BoTPA attains RI-ASRs ranging from 13.3% to 94.7% in the presence of the Krum and Multi-Krum defenses, from 2.6% to 49.2% under the Median defense, and from 2.9% to 63.5% under the Flame defense.
Authors: Zitian Gao, Yihao Xiao
Abstract: In the Venture Capital (VC) industry, predicting the success of startups is challenging due to limited financial data and the need for subjective revenue forecasts. Previous methods based on time series analysis often fall short as they fail to incorporate crucial inter-company relationships such as competition and collaboration. To fill the gap, this paper aims to introduce a novel approach using GraphRAG augmented time series model. With GraphRAG, time series predictive methods are enhanced by integrating these vital relationships into the analysis framework, allowing for a more dynamic understanding of the startup ecosystem in venture capital. Our experimental results demonstrate that our model significantly outperforms previous models in startup success predictions.
Authors: Ruixuan Liu, Alan Chen, Weiye Zhao, Changliu Liu
Abstract: Combinatorial assembly uses standardized unit primitives to build objects that satisfy user specifications. This paper studies assembly sequence planning (ASP) for physical combinatorial assembly. Given the shape of the desired object, the goal is to find a sequence of actions for placing unit primitives to build the target object. In particular, we aim to ensure the planned assembly sequence is physically executable. However, ASP for combinatorial assembly is particularly challenging due to its combinatorial nature. To address the challenge, we employ deep reinforcement learning to learn a construction policy for placing unit primitives sequentially to build the desired object. Specifically, we design an online physics-aware action mask that filters out invalid actions, which effectively guides policy learning and ensures violation-free deployment. In the end, we apply the proposed method to Lego assembly with more than 250 3D structures. The experiment results demonstrate that the proposed method plans physically valid assembly sequences to build all structures, achieving a $100\%$ success rate, whereas the best comparable baseline fails more than $40$ structures. Our implementation is available at \url{https://github.com/intelligent-control-lab/PhysicsAwareCombinatorialASP}.
URLs: https://github.com/intelligent-control-lab/PhysicsAwareCombinatorialASP
Authors: Han-Jun Choi, Hyunsung Kim, Minho Lee, Minchul Jeong, Chang-Jo Kim, Jinsung Yoon, Sang-Ki Ko
Abstract: Multi-agent trajectory data collected from domains such as team sports often suffer from missing values due to various factors. While many imputation methods have been proposed for spatiotemporal data, they are not well-suited for multi-agent sports scenarios where player movements are highly dynamic and inter-agent interactions continuously evolve. To address these challenges, we propose MIDAS (Multi-agent Imputer with Derivative-Accumulating Self-ensemble), a framework that imputes multi-agent trajectories with high accuracy and physical plausibility. It jointly predicts positions, velocities, and accelerations through a Set Transformer-based neural network and generates alternative estimates by recursively accumulating predicted velocity and acceleration values. These predictions are then combined using a learnable weighted ensemble to produce final imputed trajectories. Experiments on three sports datasets demonstrate that MIDAS significantly outperforms existing baselines in both positional accuracy and physical plausibility. Lastly, we showcase use cases of MIDAS, such as approximating total distance and pass success probability, to highlight its applicability to practical downstream tasks that require complete tracking data.
Authors: Erzhi Liu, Jerry Yao-Chieh Hu, Alex Reneau, Zhao Song, Han Liu
Abstract: We introduce a refined differentially private (DP) data structure for kernel density estimation (KDE), offering not only improved privacy-utility tradeoff but also better efficiency over prior results. Specifically, we study the mathematical problem: given a similarity function $f$ (or DP KDE) and a private dataset $X \subset \mathbb{R}^d$, our goal is to preprocess $X$ so that for any query $y\in\mathbb{R}^d$, we approximate $\sum_{x \in X} f(x, y)$ in a differentially private fashion. The best previous algorithm for $f(x,y) =\| x - y \|_1$ is the node-contaminated balanced binary tree by [Backurs, Lin, Mahabadi, Silwal, and Tarnawski, ICLR 2024]. Their algorithm requires $O(nd)$ space and time for preprocessing with $n=|X|$. For any query point, the query time is $d \log n$, with an error guarantee of $(1+\alpha)$-approximation and $\epsilon^{-1} \alpha^{-0.5} d^{1.5} R \log^{1.5} n$. In this paper, we improve the best previous result [Backurs, Lin, Mahabadi, Silwal, and Tarnawski, ICLR 2024] in three aspects: - We reduce query time by a factor of $\alpha^{-1} \log n$. - We improve the approximation ratio from $\alpha$ to 1. - We reduce the error dependence by a factor of $\alpha^{-0.5}$. From a technical perspective, our method of constructing the search tree differs from previous work [Backurs, Lin, Mahabadi, Silwal, and Tarnawski, ICLR 2024]. In prior work, for each query, the answer is split into $\alpha^{-1} \log n$ numbers, each derived from the summation of $\log n$ values in interval tree countings. In contrast, we construct the tree differently, splitting the answer into $\log n$ numbers, where each is a smart combination of two distance values, two counting values, and $y$ itself. We believe our tree structure may be of independent interest.
Authors: Haoyu Zhang, Raghavendra Ramachandra, Kiran Raja, Christoph Busch
Abstract: Face morphing attack detection (MAD) algorithms have become essential to overcome the vulnerability of face recognition systems. To solve the lack of large-scale and public-available datasets due to privacy concerns and restrictions, in this work we propose a new method to generate a synthetic face morphing dataset with 2450 identities and more than 100k morphs. The proposed synthetic face morphing dataset is unique for its high-quality samples, different types of morphing algorithms, and the generalization for both single and differential morphing attack detection algorithms. For experiments, we apply face image quality assessment and vulnerability analysis to evaluate the proposed synthetic face morphing dataset from the perspective of biometric sample quality and morphing attack potential on face recognition systems. The results are benchmarked with an existing SOTA synthetic dataset and a representative non-synthetic and indicate improvement compared with the SOTA. Additionally, we design different protocols and study the applicability of using the proposed synthetic dataset on training morphing attack detection algorithms.
Authors: Ankit Pratap Singh, Namrata Vaswani
Abstract: This letter studies the AltGDmin algorithm for solving the noisy low rank column-wise sensing (LRCS) problem. Our sample complexity guarantee improves upon the best existing one by a factor $\max(r, \log(1/\epsilon))/r$ where $r$ is the rank of the unknown matrix and $\epsilon$ is the final desired accuracy. A second contribution of this work is a detailed comparison of guarantees from all work that studies the exact same mathematical problem as LRCS, but refers to it by different names.
Authors: Arijit Khan, Tianxing Wu, Xi Chen
Abstract: The unification of large language models (LLMs) and knowledge graphs (KGs) has emerged as a hot topic. At the LLM+KG'24 workshop, held in conjunction with VLDB 2024 in Guangzhou, China, one of the key themes explored was important data management challenges and opportunities due to the effective interaction between LLMs and KGs. This report outlines the major directions and approaches presented by various speakers during the LLM+KG'24 workshop.
Authors: Rohit Gandikota, Sheridan Feucht, Samuel Marks, David Bau
Abstract: In this work, we propose Erasure of Language Memory (ELM), an approach for concept-level unlearning built on the principle of matching the distribution defined by an introspective classifier. Our key insight is that effective unlearning should leverage the model's ability to evaluate its own knowledge, using the model itself as a classifier to identify and reduce the likelihood of generating content related to undesired concepts. ELM applies this framework to create targeted low-rank updates that reduce generation probabilities for concept-specific content while preserving the model's broader capabilities. We demonstrate ELM's efficacy on biosecurity, cybersecurity, and literary domain erasure tasks. Comparative analysis shows that ELM achieves superior performance across key metrics, including near-random scores on erased topic assessments, maintained coherence in text generation, preserved accuracy on unrelated benchmarks, and robustness under adversarial attacks. Our code, data, and trained models are available at https://elm.baulab.info
URLs: https://elm.baulab.info
Authors: Mateus Karvat, Sidney Givigi
Abstract: Adverse weather conditions pose a significant challenge to the widespread adoption of Autonomous Vehicles (AVs) by impacting sensors like LiDARs and cameras. Even though Collaborative Perception (CP) improves AV perception in difficult conditions, existing CP datasets lack adverse weather conditions. To address this, we introduce Adver-City, the first open-source synthetic CP dataset focused on adverse weather conditions. Simulated in CARLA with OpenCDA, it contains over 24 thousand frames, over 890 thousand annotations, and 110 unique scenarios across six different weather conditions: clear weather, soft rain, heavy rain, fog, foggy heavy rain and, for the first time in a synthetic CP dataset, glare. It has six object categories including pedestrians and cyclists, and uses data from vehicles and roadside units featuring LiDARs, RGB and semantic segmentation cameras, GNSS, and IMUs. Its scenarios, based on real crash reports, depict the most relevant road configurations for adverse weather and poor visibility conditions, varying in object density, with both dense and sparse scenes, allowing for novel testing conditions of CP models. Benchmarks run on the dataset show that weather conditions created challenging conditions for perception models, with CoBEVT scoring 58.30/52.44/38.90 (AP@30/50/70). The dataset, code and documentation are available at https://labs.cs.queensu.ca/quarrg/datasets/adver-city/.
URLs: https://labs.cs.queensu.ca/quarrg/datasets/adver-city/.
Authors: Emiel Hoogeboom, Thomas Mensink, Jonathan Heek, Kay Lamerigts, Ruiqi Gao, Tim Salimans
Abstract: Latent diffusion models have become the popular choice for scaling up diffusion models for high resolution image synthesis. Compared to pixel-space models that are trained end-to-end, latent models are perceived to be more efficient and to produce higher image quality at high resolution. Here we challenge these notions, and show that pixel-space models can be very competitive to latent models both in quality and efficiency, achieving 1.5 FID on ImageNet512 and new SOTA results on ImageNet128, ImageNet256 and Kinetics600. We present a simple recipe for scaling end-to-end pixel-space diffusion models to high resolutions. 1: Use the sigmoid loss-weighting (Kingma & Gao, 2023) with our prescribed hyper-parameters. 2: Use our simplified memory-efficient architecture with fewer skip-connections. 3: Scale the model to favor processing the image at a high resolution with fewer parameters, rather than using more parameters at a lower resolution. Combining these with guidance intervals, we obtain a family of pixel-space diffusion models we call Simpler Diffusion (SiD2).
Authors: Shanu Kumar, Akhila Yesantarao Venkata, Shubhanshu Khandelwal, Bishal Santra, Parag Agrawal, Manish Gupta
Abstract: Prompt optimization is essential for effective utilization of large language models (LLMs) across diverse tasks. While existing optimization methods are effective in optimizing short prompts, they struggle with longer, more complex ones, often risking information loss and being sensitive to small perturbations. To address these challenges, we propose SCULPT (Systematic Tuning of Long Prompts), a framework that treats prompt optimization as a hierarchical tree refinement problem. SCULPT represents prompts as tree structures, enabling targeted modifications while preserving contextual integrity. It employs a Critic-Actor framework that generates reflections and applies actions to refine the prompt. Evaluations demonstrate SCULPT's effectiveness on long prompts, its robustness to adversarial perturbations, and its ability to generate high-performing prompts even without any initial human-written prompt. Compared to existing state of the art methods, SCULPT consistently improves LLM performance by preserving essential task information while applying structured refinements. Both qualitative and quantitative analyses show that SCULPT produces more stable and interpretable prompt modifications, ensuring better generalization across tasks.
Authors: Hesam Hosseini, Ghazal Hosseini Mighan, Amirabbas Afzali, Sajjad Amini, Amir Houmansadr
Abstract: Transformers have revolutionized Computer Vision (CV) through self-attention mechanisms. However, their complexity makes latent token representations difficult to interpret. We introduce ULTra, a framework for interpreting Transformer embeddings and uncovering meaningful semantic patterns within them. ULTra enables unsupervised semantic segmentation using pre-trained models without requiring fine-tuning. Additionally, we propose a self-supervised training approach that refines segmentation performance by learning an external transformation matrix without modifying the underlying model. Our method achieves state-of-the-art performance in unsupervised semantic segmentation, outperforming existing segmentation methods. Furthermore, we validate ULTra for model interpretation on both synthetic and real-world scenarios, including Object Selection and interpretable text summarization using LLMs, demonstrating its broad applicability in explaining the semantic structure of latent token representations.
Authors: Yiming Zhang, Zhuokai Zhao, Zhaorun Chen, Zenghui Ding, Xianjun Yang, Yining Sun
Abstract: Recent advancements in multimodal large language models (MLLMs) have opened new avenues for video understanding. However, achieving high fidelity in zero-shot video tasks remains challenging. Traditional video processing methods rely heavily on fine-tuning to capture nuanced spatial-temporal details, which incurs significant data and computation costs. In contrast, training-free approaches, though efficient, often lack robustness in preserving context-rich features across complex video content. To this end, we propose DYTO, a novel dynamic token merging framework for zero-shot video understanding that adaptively optimizes token efficiency while preserving crucial scene details. DYTO integrates a hierarchical frame selection and a bipartite token merging strategy to dynamically cluster key frames and selectively compress token sequences, striking a balance between computational efficiency with semantic richness. Extensive experiments across multiple benchmarks demonstrate the effectiveness of DYTO, achieving superior performance compared to both fine-tuned and training-free methods and setting a new state-of-the-art for zero-shot video understanding.
Authors: Hyelin Nam, Jaemin Kim, Dohun Lee, Jong Chul Ye
Abstract: While text-to-video diffusion models have made significant strides, many still face challenges in generating videos with temporal consistency. Within diffusion frameworks, guidance techniques have proven effective in enhancing output quality during inference; however, applying these methods to video diffusion models introduces additional complexity of handling computations across entire sequences. To address this, we propose a novel framework called MotionPrompt that guides the video generation process via optical flow. Specifically, we train a discriminator to distinguish optical flow between random pairs of frames from real videos and generated ones. Given that prompts can influence the entire video, we optimize learnable token embeddings during reverse sampling steps by using gradients from a trained discriminator applied to random frame pairs. This approach allows our method to generate visually coherent video sequences that closely reflect natural motion dynamics, without compromising the fidelity of the generated content. We demonstrate the effectiveness of our approach across various models.
Authors: Zhixuan Liang, Yao Mu, Yixiao Wang, Tianxing Chen, Wenqi Shao, Wei Zhan, Masayoshi Tomizuka, Ping Luo, Mingyu Ding
Abstract: Dexterous manipulation with contact-rich interactions is crucial for advanced robotics. While recent diffusion-based planning approaches show promise for simple manipulation tasks, they often produce unrealistic ghost states (e.g., the object automatically moves without hand contact) or lack adaptability when handling complex sequential interactions. In this work, we introduce DexDiffuser, an interaction-aware diffusion planning framework for adaptive dexterous manipulation. DexDiffuser models joint state-action dynamics through a dual-phase diffusion process which consists of pre-interaction contact alignment and post-contact goal-directed control, enabling goal-adaptive generalizable dexterous manipulation. Additionally, we incorporate dynamics model-based dual guidance and leverage large language models for automated guidance function generation, enhancing generalizability for physical interactions and facilitating diverse goal adaptation through language cues. Experiments on physical interaction tasks such as door opening, pen and block re-orientation, object relocation, and hammer striking demonstrate DexDiffuser's effectiveness on goals outside training distributions, achieving over twice the average success rate (59.2% vs. 29.5%) compared to existing methods. Our framework achieves an average of 70.7% success rate on goal adaptive dexterous tasks, highlighting its robustness and flexibility in contact-rich manipulation.
Authors: Ahc\`ene Boubekki, Samuel G. Fadel, Sebastian Mair
Abstract: This paper introduces Neuro-Activated Vision Explanations (NAVE), a method for extracting and visualizing the internal representations of vision model encoders. By clustering feature activations, NAVE provides insights into learned semantics without fine-tuning. Using object localization, we show that NAVE's concepts align with image semantics. Through extensive experiments, we analyze the impact of training strategies and architectures on encoder representation capabilities. Additionally, we apply NAVE to study training artifacts in vision transformers and reveal how weak training strategies and spurious correlations degrade model performance. Our findings establish NAVE as a valuable tool for post-hoc model inspection and improving transparency in vision models.
Authors: Xinjie Cui, Yuezun Li, Ao Luo, Jiaran Zhou, Junyu Dong
Abstract: We describe the Forensics Adapter, an adapter network designed to transform CLIP into an effective and generalizable face forgery detector. Although CLIP is highly versatile, adapting it for face forgery detection is non-trivial as forgery-related knowledge is entangled with a wide range of unrelated knowledge. Existing methods treat CLIP merely as a feature extractor, lacking task-specific adaptation, which limits their effectiveness. To address this, we introduce an adapter to learn face forgery traces -- the blending boundaries unique to forged faces, guided by task-specific objectives. Then we enhance the CLIP visual tokens with a dedicated interaction strategy that communicates knowledge across CLIP and the adapter. Since the adapter is alongside CLIP, its versatility is highly retained, naturally ensuring strong generalizability in face forgery detection. With only 5.7M trainable parameters, our method achieves a significant performance boost, improving by approximately 7% on average across five standard datasets. We believe the proposed method can serve as a baseline for future CLIP-based face forgery detection methods. The code is available at https://github.com/OUC-VAS/ForensicsAdapter.
Authors: Friedhelm Hamann, Daniel Gehrig, Filbert Febryanto, Kostas Daniilidis, Guillermo Gallego
Abstract: Tracking any point (TAP) recently shifted the motion estimation paradigm from focusing on individual salient points with local templates to tracking arbitrary points with global image contexts. However, while research has mostly focused on driving the accuracy of models in nominal settings, addressing scenarios with difficult lighting conditions and high-speed motions remains out of reach due to the limitations of the sensor. This work addresses this challenge with the first event camera-based TAP method. It leverages the high temporal resolution and high dynamic range of event cameras for robust high-speed tracking, and the global contexts in TAP methods to handle asynchronous and sparse event measurements. We further extend the TAP framework to handle event feature variations induced by motion -- thereby addressing an open challenge in purely event-based tracking -- with a novel feature-alignment loss which ensures the learning of motion-robust features. Our method is trained with data from a new data generation pipeline and systematically ablated across all design decisions. Our method shows strong cross-dataset generalization and performs 136% better on the average Jaccard metric than the baselines. Moreover, on an established feature tracking benchmark, it achieves a 20% improvement over the previous best event-only method and even surpasses the previous best events-and-frames method by 4.1%. Our code is available at https://github.com/tub-rip/ETAP
Authors: Gin\'es Carreto Pic\'on, Illia Oleksiienko, Lukas Hedegaard, Arian Bakhtiarnia, Alexandros Iosifidis
Abstract: Transformers are widely used for their ability to capture data relations in sequence processing, with great success for a wide range of static tasks. However, the computational and memory footprint of their main component, i.e., the Scaled Dot-product Attention, is commonly overlooked. This makes their adoption in applications involving stream data processing with constraints in response latency, computational and memory resources infeasible. Some works have proposed methods to lower the computational cost of Transformers, i.e. low-rank approximations, sparsity in attention, and efficient formulations for Continual Inference. In this paper, we introduce a new formulation of the Scaled Dot-product Attention based on the Nystr\"om approximation that is suitable for Continual Inference. In experiments on Online Audio Classification and Online Action Detection tasks, the proposed Continual Scaled Dot-product Attention can lower the number of operations by up to three orders of magnitude compared to the original Transformers while retaining the predictive performance of competing models.
Authors: Surya Dheeshjith, Adam Subel, Alistair Adcroft, Julius Busecke, Carlos Fernandez-Granda, Shubham Gupta, Laure Zanna
Abstract: AI emulators for forecasting have emerged as powerful tools that can outperform conventional numerical predictions. The next frontier is to build emulators for long climate simulations with skill across a range of spatiotemporal scales, a particularly important goal for the ocean. Our work builds a skillful global emulator of the ocean component of a state-of-the-art climate model. We emulate key ocean variables, sea surface height, horizontal velocities, temperature, and salinity, across their full depth. We use a modified ConvNeXt UNet architecture trained on multidepth levels of ocean data. We show that the ocean emulator - Samudra - which exhibits no drift relative to the truth, can reproduce the depth structure of ocean variables and their interannual variability. Samudra is stable for centuries and 150 times faster than the original ocean model. Samudra struggles to capture the correct magnitude of the forcing trends and simultaneously remain stable, requiring further work.
Authors: Wenyu Bo (Department of Statistics University of Washington Seattle, WA), Marina Meil\u{a} (Department of Statistics University of Washington Seattle, WA)
Abstract: Under a set of assumptions on a family of submanifolds $\subset {\mathbb R}^D$, we derive a series of geometric properties that remain valid after finite-dimensional and almost isometric Diffusion Maps (DM), including almost uniform density, finite polynomial approximation and reach. Leveraging these properties, we establish rigorous bounds on the embedding errors introduced by the DM algorithm is $O\left((\frac{\log n}{n})^{\frac{1}{8d+16}}\right)$. Furthermore, we quantify the error between the estimated tangent spaces and the true tangent spaces over the submanifolds after the DM embedding, $\sup_{P\in \mathcal{P}}\mathbb{E}_{P^{\otimes \tilde{n}}} \max_{1\leq j \angle (T_{Y_{\varphi(M),j}}\varphi(M),\hat{T}_j)\leq \tilde{n}} \leq C \left(\frac{\log n }{n}\right)^\frac{k-1}{(8d+16)k}$, which providing a precise characterization of the geometric accuracy of the embeddings. These results offer a solid theoretical foundation for understanding the performance and reliability of DM in practical applications.
Authors: Chaozheng Wen, Jingwen Tong, Yingdong Hu, Zehong Lin, Jun Zhang
Abstract: Wireless channel modeling plays a pivotal role in designing, analyzing, and optimizing wireless communication systems. Nevertheless, developing an effective channel modeling approach has been a long-standing challenge. This issue has been escalated due to denser network deployment, larger antenna arrays, and broader bandwidth in next-generation networks. To address this challenge, we put forth WRF-GS, a novel framework for channel modeling based on wireless radiation field (WRF) reconstruction using 3D Gaussian splatting (3D-GS). WRF-GS employs 3D Gaussian primitives and neural networks to capture the interactions between the environment and radio signals, enabling efficient WRF reconstruction and visualization of the propagation characteristics. The reconstructed WRF can then be used to synthesize the spatial spectrum for comprehensive wireless channel characterization. While WRF-GS demonstrates remarkable effectiveness, it faces limitations in capturing high-frequency signal variations caused by complex multipath effects. To overcome these limitations, we propose WRF-GS+, an enhanced framework that integrates electromagnetic wave physics into the neural network design. WRF-GS+ leverages deformable 3D Gaussians to model both static and dynamic components of the WRF, significantly improving its ability to characterize signal variations. In addition, WRF-GS+ enhances the splatting process by simplifying the 3D-GS modeling process and improving computational efficiency. Experimental results demonstrate that both WRF-GS and WRF-GS+ outperform baselines for spatial spectrum synthesis, including ray tracing and other deep-learning approaches. Notably, WRF-GS+ achieves state-of-the-art performance in the received signal strength indication (RSSI) and channel state information (CSI) prediction tasks, surpassing existing methods by more than 0.7 dB and 3.36 dB, respectively.
Authors: Kevin Murphy
Abstract: This manuscript gives a big-picture, up-to-date overview of the field of (deep) reinforcement learning and sequential decision making, covering value-based method, policy-gradient methods, model-based methods, and various other topics (e.g., multi-agent RL, RL+LLMs, and RL+inference).
Authors: Ximing Xing, Juncheng Hu, Jing Zhang, Dong Xu, Qian Yu
Abstract: In this work, we introduce SVGFusion, a Text-to-SVG model capable of scaling to real-world SVG data without relying on text-based discrete language models or prolonged Score Distillation Sampling (SDS) optimization. The core idea of SVGFusion is to utilize a popular Text-to-Image framework to learn a continuous latent space for vector graphics. Specifically, SVGFusion comprises two key modules: a Vector-Pixel Fusion Variational Autoencoder (VP-VAE) and a Vector Space Diffusion Transformer (VS-DiT). The VP-VAE processes both SVG codes and their corresponding rasterizations to learn a continuous latent space, while the VS-DiT generates latent codes within this space based on the input text prompt. Building on the VP-VAE, we propose a novel rendering sequence modeling strategy which enables the learned latent space to capture the inherent creation logic of SVGs. This allows the model to generate SVGs with higher visual quality and more logical construction, while systematically avoiding occlusion in complex graphic compositions. Additionally, the scalability of SVGFusion can be continuously enhanced by adding more VS-DiT blocks. To effectively train and evaluate SVGFusion, we construct SVGX-Dataset, a large-scale, high-quality SVG dataset that addresses the scarcity of high-quality vector data. Extensive experiments demonstrate the superiority of SVGFusion over existing SVG generation methods, establishing a new framework for SVG content creation. Code, model, and data will be released at: https://ximinng.github.io/SVGFusionProject/
Authors: Benjamin Laufer, Manish Raghavan, Solon Barocas
Abstract: Disparate impact doctrine offers an important legal apparatus for targeting discriminatory data-driven algorithmic decisions. A recent body of work has focused on conceptualizing one particular construct from this doctrine: the less discriminatory alternative, an alternative policy that reduces disparities while meeting the same business needs of a status quo or baseline policy. However, attempts to operationalize this construct in the algorithmic setting must grapple with some thorny challenges and ambiguities. In this paper, we attempt to raise and resolve important questions about less discriminatory algorithms (LDAs). How should we formally define LDAs, and how does this interact with different societal goals they might serve? And how feasible is it for firms or plaintiffs to computationally search for candidate LDAs? We find that formal LDA definitions face fundamental challenges when they attempt to evaluate and compare predictive models in the absence of held-out data. As a result, we argue that LDA definitions cannot be purely quantitative, and must rely on standards of "reasonableness." We then identify both mathematical and computational constraints on firms' ability to efficiently conduct a proactive search for LDAs, but we provide evidence that these limits are "weak" in a formal sense. By defining LDAs formally, we put forward a framework in which both firms and plaintiffs can search for alternative models that comport with societal goals.
Authors: Brendan Mallery, James M. Murphy, Shuchin Aeron
Abstract: We consider synthesis and analysis of probability measures using the entropy-regularized Wasserstein-2 cost and its unbiased version, the Sinkhorn divergence. The synthesis problem consists of computing the barycenter, with respect to these costs, of reference measures given a set of coefficients belonging to the simplex. The analysis problem consists of finding the coefficients for the closest barycenter in the Wasserstein-2 distance to a given measure. Under the weakest assumptions on the measures thus far in the literature, we compute the derivative of the entropy-regularized Wasserstein-2 cost. We leverage this to establish a characterization of barycenters with respect to the entropy-regularized Wasserstein-2 cost as solutions that correspond to a fixed point of an average of the entropy-regularized displacement maps. This characterization yields a finite-dimensional, convex, quadratic program for solving the analysis problem when the measure being analyzed is a barycenter with respect to the entropy-regularized Wasserstein-2 cost. We show that these coefficients, as well as the value of the barycenter functional, can be estimated from samples with dimension-independent rates of convergence, and that barycentric coefficients are stable with respect to perturbations in the Wasserstein-2 metric. We employ the barycentric coefficients as features for classification of corrupted point cloud data, and show that compared to neural network baselines, our approach is more efficient in small training data regimes.
Authors: Hayder Tirmazi
Abstract: The Classical Bloom Filter (CBF) is a class of Probabilistic Data Structures (PDS) for handling Approximate Query Membership (AMQ). The Learned Bloom Filter (LBF) is a recently proposed class of PDS that combines the Classical Bloom Filter with a Learning Model while preserving the Bloom Filter's one-sided error guarantees. Bloom Filters have been used in settings where inputs are sensitive and need to be private in the presence of an adversary with access to the Bloom Filter through an API or in the presence of an adversary who has access to the internal state of the Bloom Filter. This paper conducts a rigorous differential privacy-based analysis for the Bloom Filter. We propose constructions that satisfy differential privacy and asymmetric differential privacy. This is also the first work that analyses and addresses the privacy of the Learned Bloom Filter under any rigorous model, which is an open problem.
Authors: Sina Sharifi, Nazanin Abolfazli, Erfan Yazdandoost Hamedani, Mahyar Fazlyab
Abstract: Bilevel optimization is a key framework in hierarchical decision-making, where one problem is embedded within the constraints of another. In this work, we propose a control-theoretic approach to solving bilevel optimization problems. Our method consists of two components: a gradient flow mechanism to minimize the upper-level objective and a safety filter to enforce the constraints imposed by the lower-level problem. Together, these components form a safe gradient flow that solves the bilevel problem in a single loop. To improve scalability with respect to the lower-level problem's dimensions, we introduce a relaxed formulation and design a compact variant of the safe gradient flow. This variant minimizes the upper-level objective while ensuring the lower-level decision variable remains within a user-defined suboptimality. Using Lyapunov analysis, we establish convergence guarantees for the dynamics, proving that they converge to a neighborhood of the optimal solution. Numerical experiments further validate the effectiveness of the proposed approaches. Our contributions provide both theoretical insights and practical tools for efficiently solving bilevel optimization problems.
Authors: Maja Pavlovic
Abstract: To be considered reliable, a model must be calibrated so that its confidence in each decision closely reflects its true outcome. In this blogpost we'll take a look at the most commonly used definition for calibration and then dive into a frequently used evaluation measure for model calibration. We'll then cover some of the drawbacks of this measure and how these surfaced the need for additional notions of calibration, which require their own new evaluation measures. This post is not intended to be an in-depth dissection of all works on calibration, nor does it focus on how to calibrate models. Instead, it is meant to provide a gentle introduction to the different notions and their evaluation measures as well as to re-highlight some issues with a measure that is still widely used to evaluate calibration.
Authors: Kwanghee Choi, Eunjung Yeo, Kalvin Chang, Shinji Watanabe, David Mortensen
Abstract: Allophony refers to the variation in the phonetic realization of a phoneme based on its phonetic environment. Modeling allophones is crucial for atypical pronunciation assessment, which involves distinguishing atypical from typical pronunciations. However, recent phoneme classifier-based approaches often simplify this by treating various realizations as a single phoneme, bypassing the complexity of modeling allophonic variation. Motivated by the acoustic modeling capabilities of frozen self-supervised speech model (S3M) features, we propose MixGoP, a novel approach that leverages Gaussian mixture models to model phoneme distributions with multiple subclusters. Our experiments show that MixGoP achieves state-of-the-art performance across four out of five datasets, including dysarthric and non-native speech. Our analysis further suggests that S3M features capture allophonic variation more effectively than MFCCs and Mel spectrograms, highlighting the benefits of integrating MixGoP with S3M features.
Authors: Ao Li, Wei Fang, Hongbo Zhao, Le Lu, Ge Yang, Minfeng Xu
Abstract: In applications of diffusion models, controllable generation is of practical significance, but is also challenging. Current methods for controllable generation primarily focus on modifying the score function of diffusion models, while Mean Reverting (MR) Diffusion directly modifies the structure of the stochastic differential equation (SDE), making the incorporation of image conditions simpler and more natural. However, current training-free fast samplers are not directly applicable to MR Diffusion. And thus MR Diffusion requires hundreds of NFEs (number of function evaluations) to obtain high-quality samples. In this paper, we propose a new algorithm named MaRS (MR Sampler) to reduce the sampling NFEs of MR Diffusion. We solve the reverse-time SDE and the probability flow ordinary differential equation (PF-ODE) associated with MR Diffusion, and derive semi-analytical solutions. The solutions consist of an analytical function and an integral parameterized by a neural network. Based on this solution, we can generate high-quality samples in fewer steps. Our approach does not require training and supports all mainstream parameterizations, including noise prediction, data prediction and velocity prediction. Extensive experiments demonstrate that MR Sampler maintains high sampling quality with a speedup of 10 to 20 times across ten different image restoration tasks. Our algorithm accelerates the sampling procedure of MR Diffusion, making it more practical in controllable generation.
Authors: Xiaoxia Xu, Xidong Mu, Yuanwei Liu, Arumugam Nallanathan
Abstract: A novel pinching antenna system (PASS)-enabled downlink multi-user multiple-input single-output (MISO) framework is proposed. PASS consists of multiple waveguides spanning over thousands of wavelength, which equip numerous low-cost dielectric particles, named pinching antennas (PAs), to radiate signals into free space. The positions of PAs can be reconfigured to change both the large-scale path losses and phases of signals, thus facilitating the novel pinching beamforming design. A sum rate maximization problem is formulated, which jointly optimizes the transmit and pinching beamforming to adaptively achieve constructive signal enhancement and destructive interference mitigation. To solve this highly coupled and nonconvex problem, both optimization-based and learning-based methods are proposed. 1) For the optimization-based method, a majorization-minimization and penalty dual decomposition (MM-PDD) algorithm is developed, which handles the nonconvex complex exponential component using a Lipschitz surrogate function and then invokes PDD for problem decoupling. 2) For the learning-based method, a novel Karush-Kuhn-Tucker (KKT)-guided dual learning (KDL) approach is proposed, which enables KKT solutions to be reconstructed in a data-driven manner by learning dual variables. Following this idea, a KDL-Tranformer algorithm is developed, which captures both inter-PA/inter-user dependencies and channel-state-information (CSI)-beamforming dependencies by attention mechanisms. Simulation results demonstrate that: i) The proposed PASS framework significantly outperforms conventional massive multiple input multiple output (MIMO) system even with a few PAs. ii) The proposed KDL-Transformer can improve over 30% system performance than MM-PDD algorithm, while achieving a millisecond-level response on modern GPUs.
Authors: Tommaso Mencattini, Adrian Robert Minut, Donato Crisostomi, Andrea Santilli, Emanuele Rodol\`a
Abstract: Evolutionary model merging enables the creation of high-performing multi-task models but remains computationally prohibitive for consumer hardware. We introduce MERGE$^3$, an efficient framework that makes evolutionary merging feasible on a single GPU by reducing fitness computation costs 50$\times$ while preserving performance. MERGE$^3$ achieves this by Extracting a reduced dataset for evaluation, Estimating model abilities using Item Response Theory (IRT), and Evolving optimal merges via IRT-based performance estimators. Our method enables state-of-the-art multilingual and cross-lingual merging, transferring knowledge across languages with significantly lower computational overhead. We provide theoretical guarantees and an open-source library, democratizing high-quality model merging.
Authors: Fengwei Teng, Zhaoyang Yu, Quan Shi, Jiayi Zhang, Chenglin Wu, Yuyu Luo
Abstract: Large Language Models (LLMs) achieve superior performance through training-time scaling, and test-time scaling further enhances their capabilities by conducting effective reasoning during inference. However, as the scale of reasoning increases, existing test-time scaling methods suffer from accumulated historical information, which not only wastes computational resources but also interferes with effective reasoning. To address this issue, we observe that complex reasoning can be achieved by solving a series of independent and self-contained subquestions. These subquestions are essentially \textit{atomic questions}, exhibiting the memoryless property similar to Markov processes. Based on this observation, we propose Atom of Thoughts (\our), where each state transition consists of decomposing the current question into a dependency-based directed acyclic graph and contracting its subquestions, forming a simplified question that maintains answer equivalence with the original problem. This answer preservation enables the iterative \textit{decomposition-contraction} process to naturally form a meaningful Markov reasoning process. Furthermore, these atomic states can be seamlessly integrated into existing test-time scaling methods, enabling \our to serve as a plug-in enhancement for improving reasoning capabilities. Experiments across six benchmarks demonstrate the effectiveness of \our both as a standalone framework and a plug-in enhancement. Notably, on HotpotQA, when applied to gpt-4o-mini, \our achieves an \textbf{80.6\%} F1 score, surpassing o3-mini by \textbf{3.4\%} and DeepSeek-R1 by \textbf{10.6\%}. The code is available at \href{https://github.com/qixucen/atom}{https://github.com/qixucen/atom}.
URLs: https://github.com/qixucen/atom, https://github.com/qixucen/atom
Authors: Tianqing Wang, Xun Xue, Guoliang Li, Yong Wang
Abstract: In this demonstration, we present AnDB, an AI-native database that supports traditional OLTP workloads and innovative AI-driven tasks, enabling unified semantic analysis across structured and unstructured data. While structured data analytics is mature, challenges remain in bridging the semantic gap between user queries and unstructured data. AnDB addresses these issues by leveraging cutting-edge AI-native technologies, allowing users to perform semantic queries using intuitive SQL-like statements without requiring AI expertise. This approach eliminates the ambiguity of traditional text-to-SQL systems and provides a seamless end-to-end optimization for analyzing all data types. AnDB automates query processing by generating multiple execution plans and selecting the optimal one through its optimizer, which balances accuracy, execution time, and financial cost based on user policies and internal optimizing mechanisms. AnDB future-proofs data management infrastructure, empowering users to effectively and efficiently harness the full potential of all kinds of data without starting from scratch.
Authors: Dongkun Zhang, Jiaming Liang, Ke Guo, Sha Lu, Qi Wang, Rong Xiong, Zhenwei Miao, Yue Wang
Abstract: Trajectory planning is vital for autonomous driving, ensuring safe and efficient navigation in complex environments. While recent learning-based methods, particularly reinforcement learning (RL), have shown promise in specific scenarios, RL planners struggle with training inefficiencies and managing large-scale, real-world driving scenarios. In this paper, we introduce \textbf{CarPlanner}, a \textbf{C}onsistent \textbf{a}uto-\textbf{r}egressive \textbf{Planner} that uses RL to generate multi-modal trajectories. The auto-regressive structure enables efficient large-scale RL training, while the incorporation of consistency ensures stable policy learning by maintaining coherent temporal consistency across time steps. Moreover, CarPlanner employs a generation-selection framework with an expert-guided reward function and an invariant-view module, simplifying RL training and enhancing policy performance. Extensive analysis demonstrates that our proposed RL framework effectively addresses the challenges of training efficiency and performance enhancement, positioning CarPlanner as a promising solution for trajectory planning in autonomous driving. To the best of our knowledge, we are the first to demonstrate that the RL-based planner can surpass both IL- and rule-based state-of-the-arts (SOTAs) on the challenging large-scale real-world dataset nuPlan. Our proposed CarPlanner surpasses RL-, IL-, and rule-based SOTA approaches within this demanding dataset.
Authors: Farouk Mokhtar, Joosep Pata, Dolores Garcia, Eric Wulff, Mengke Zhang, Michael Kagan, Javier Duarte
Abstract: We demonstrate transfer learning capabilities in a machine-learned algorithm trained for particle-flow reconstruction in high energy particle colliders. This paper presents a cross-detector fine-tuning study, where we initially pre-train the model on a large full simulation dataset from one detector design, and subsequently fine-tune the model on a sample with a different collider and detector design. Specifically, we use the Compact Linear Collider detector (CLICdet) model for the initial training set, and demonstrate successful knowledge transfer to the CLIC-like detector (CLD) proposed for the Future Circular Collider in electron-positron mode (FCC-ee). We show that with an order of magnitude less samples from the second dataset, we can achieve the same performance as a costly training from scratch, across particle-level and event-level performance metrics, including jet and missing transverse momentum resolution. Furthermore, we find that the fine-tuned model achieves comparable performance to the traditional rule-based particle-flow approach on event-level metrics after training on 100,000 CLD events, whereas a model trained from scratch requires at least 1 million CLD events to achieve similar reconstruction performance. To our knowledge, this represents the first full-simulation cross-detector transfer learning study for particle-flow reconstruction. These findings offer valuable insights towards building large foundation models that can be fine-tuned across different detector designs and geometries, helping to accelerate the development cycle for new detectors and opening the door to rapid detector design and optimization using machine learning.
Authors: Zahra Zanjani Foumani, Ramin Bostanabad
Abstract: Bayesian optimization (BO) is increasingly employed in critical applications to find the optimal design with minimal cost. While BO is known for its sample efficiency, relying solely on costly high-fidelity data can still result in high costs. This is especially the case in constrained search spaces where BO must not only optimize but also ensure feasibility. A related issue in the BO literature is the lack of a systematic stopping criterion. To solve these challenges, we develop a constrained cost-aware multi-fidelity BO (CMFBO) framework whose goal is to minimize overall sampling costs by utilizing inexpensive low-fidelity sources while ensuring feasibility. In our case, the constraints can change across the data sources and may be even black-box functions. We also introduce a systematic stopping criterion that addresses the long-lasting issue associated with BO's convergence assessment. Our framework is publicly available on GitHub through the GP+ Python package and herein we validate it's efficacy on multiple benchmark problems.
Authors: Shun Liao, Paolo Di Achille, Jiang Wu, Silviu Borac, Jonathan Wang, Xin Liu, Eric Teasley, Lawrence Cai, Yuzhe Yang, Yun Liu, Daniel McDuff, Hao-Wei Su, Brent Winslow, Anupam Pathak, Shwetak Patel, James A. Taylor, Jameson K. Rogers, Ming-Zher Poh
Abstract: Resting heart rate (RHR) is an important biomarker of cardiovascular health and mortality, but tracking it longitudinally generally requires a wearable device, limiting its availability. We present PHRM, a deep learning system for passive heart rate (HR) and RHR measurements during everyday smartphone use, using facial video-based photoplethysmography. Our system was developed using 225,773 videos from 495 participants and validated on 185,970 videos from 205 participants in laboratory and free-living conditions, representing the largest validation study of its kind. Compared to reference electrocardiogram, PHRM achieved a mean absolute percentage error (MAPE) < 10% for HR measurements across three skin tone groups of light, medium and dark pigmentation; MAPE for each skin tone group was non-inferior versus the others. Daily RHR measured by PHRM had a mean absolute error < 5 bpm compared to a wearable HR tracker, and was associated with known risk factors. These results highlight the potential of smartphones to enable passive and equitable heart health monitoring.
Authors: Zhijian Zhuo, Yutao Zeng, Ya Wang, Sijun Zhang, Jian Yang, Xiaoqing Li, Xun Zhou, Jinwen Ma
Abstract: Transformers have become the de facto architecture for a wide range of machine learning tasks, particularly in large language models (LLMs). Despite their remarkable performance, challenges remain in training deep transformer networks, especially regarding the location of layer normalization. While Pre-Norm structures facilitate easier training due to their more prominent identity path, they often yield suboptimal performance compared to Post-Norm. In this paper, we propose $\textbf{HybridNorm}$, a straightforward yet effective hybrid normalization strategy that integrates the advantages of both Pre-Norm and Post-Norm approaches. Specifically, HybridNorm employs QKV normalization within the attention mechanism and Post-Norm in the feed-forward network (FFN) of each transformer block. This design not only stabilizes training but also enhances performance, particularly in the context of LLMs. Comprehensive experiments in both dense and sparse architectures show that HybridNorm consistently outperforms both Pre-Norm and Post-Norm approaches, achieving state-of-the-art results across various benchmarks. These findings highlight the potential of HybridNorm as a more stable and effective technique for improving the training and performance of deep transformer models. Code is available at https://github.com/BryceZhuo/HybridNorm.
Authors: Zebin You, Jingyang Ou, Xiaolu Zhang, Jun Hu, Jun Zhou, Chongxuan Li
Abstract: Although masked image generation models and masked diffusion models are designed with different motivations and objectives, we observe that they can be unified within a single framework. Building upon this insight, we carefully explore the design space of training and sampling, identifying key factors that contribute to both performance and efficiency. Based on the improvements observed during this exploration, we develop our model, referred to as eMIGM. Empirically, eMIGM demonstrates strong performance on ImageNet generation, as measured by Fr\'echet Inception Distance (FID). In particular, on ImageNet 256x256, with similar number of function evaluations (NFEs) and model parameters, eMIGM outperforms the seminal VAR. Moreover, as NFE and model parameters increase, eMIGM achieves performance comparable to the state-of-the-art continuous diffusion models while requiring less than 40% of the NFE. Additionally, on ImageNet 512x512, with only about 60% of the NFE, eMIGM outperforms the state-of-the-art continuous diffusion models. Code is available at https://github.com/ML-GSAI/eMIGM.
Authors: Ce Guo, Tong Zhao
Abstract: Field-Programmable Gate Arrays (FPGAs) are widely used in modern hardware design, yet writing Hardware Description Language (HDL) code for FPGA implementation remains a complex and time-consuming task. Large Language Models (LLMs) have emerged as a promising tool for HDL generation, but existing benchmarks for LLM-based code generation primarily focus on functional correctness while overlooking hardware resource usage. Furthermore, current benchmarks offer limited diversity and do not fully represent the wide range of real-world FPGA applications. To address these shortcomings, we introduce ResBench, the first resource-focused benchmark explicitly designed to distinguish between resource-optimized and inefficient LLM-generated HDL code. ResBench consists of 56 problems across 12 categories, covering applications from finite state machines to financial computing. Our open-source evaluation framework automatically tests LLMs by generating Verilog code, verifying correctness, and measuring resource usage. The experiments, which primarily analyze Lookup Table (LUT) usage, reveal significant differences among LLMs, demonstrating ResBench's capability to identify models that generate more resource-optimized FPGA designs.
Authors: Alberto Coffrini, Mohammad Amin Zadenoori, Paolo Barsocchi, Francesco Furfari, Antonino Crivello, Alessio Ferrari
Abstract: Indoor navigation presents unique challenges due to complex layouts, lack of GPS signals, and accessibility concerns. Existing solutions often struggle with real-time adaptability and user-specific needs. In this work, we explore the potential of a Large Language Model (LLM), i.e., ChatGPT, to generate natural, context-aware navigation instructions from indoor map images. We design and evaluate test cases across different real-world environments, analyzing the effectiveness of LLMs in interpreting spatial layouts, handling user constraints, and planning efficient routes. Our findings demonstrate the potential of LLMs for supporting personalized indoor navigation, with an average of 50.54% correct indications and a maximum of 77.78%. The results do not appear to depend on the complexity of the layout or the complexity of the expected path, but rather on the number of points of interest and the abundance of visual information, which negatively affect the performance.
Authors: Artem Lensky
Abstract: This study investigates EEG as a potential early biomarker by applying deep learning techniques to resting-state EEG recordings from 31 subjects (15 with PD and 16 healthy controls). EEG signals underwent preprocessing to remove tremor artifacts before classification with CNNs using wavelet-based electrode triplet images. Our analysis across different brain regions and frequency bands showed distinct spatial-spectral patterns of PD-related neural oscillations. We identified high classification accuracy (76%) using central electrodes (C3, Cz, C4) with full-spectrum 0.4-62.4 Hz analysis and 74% accuracy in right parietal regions (P8, CP6, P4) with 10-second windows. Bilateral centro-parietal regions showed strong performance (67%) in the theta band (4.0-7.79 Hz), while multiple areas demonstrated some sensitivity (65%) in the alpha band (7.8-15.59 Hz). We also observed a distinctive topographical pattern of gamma band (40-62.4 Hz) alterations specifically localized to central-parietal regions, which remained consistent across different temporal windows. In particular, we observed pronounced right-hemisphere involvement across several frequency bands. Unlike previous studies that achieved higher accuracies by potentially including tremor artifacts, our approach isolates genuine neurophysiological alterations in cortical activity. These findings suggest that specific EEG-based oscillatory patterns, especially in central and parietal regions and across multiple frequency bands, may provide diagnostic information for PD, potentially before the onset of motor symptoms.
Authors: Jingyuan Xue, Longfei Wei, Fang Sheng, Yuxin Gao, Jianfei Zhang
Abstract: The accurate prediction of RUL for lithium-ion batteries is crucial for enhancing the reliability and longevity of energy storage systems. Traditional methods for RUL prediction often struggle with issues such as data sparsity, varying battery chemistries, and the inability to capture complex degradation patterns over time. In this study, we propose a survival analysis-based framework combined with deep learning models to predict the RUL of lithium-ion batteries. Specifically, we utilize five advanced models: the Cox-type models (Cox, CoxPH, and CoxTime) and two machine-learning-based models (DeepHit and MTLR). These models address the challenges of accurate RUL estimation by transforming raw time-series battery data into survival data, including key degradation indicators such as voltage, current, and internal resistance. Advanced feature extraction techniques enhance the model's robustness in diverse real-world scenarios, including varying charging conditions and battery chemistries. Our models are tested using 10-fold cross-validation, ensuring generalizability and minimizing overfitting. Experimental results show that our survival-based framework significantly improves RUL prediction accuracy compared to traditional methods, providing a reliable tool for battery management and maintenance optimization. This study contributes to the advancement of predictive maintenance in battery technology, offering valuable insights for both researchers and industry practitioners aiming to enhance the operational lifespan of lithium-ion batteries.
Authors: Mert Bulent Sariyildiz, Philippe Weinzaepfel, Thomas Lucas, Pau de Jorge, Diane Larlus, Yannis Kalantidis
Abstract: Recent multi-teacher distillation methods have unified the encoders of multiple foundation models into a single encoder, achieving competitive performance on core vision tasks like classification, segmentation, and depth estimation. This led us to ask: Could similar success be achieved when the pool of teachers also includes vision models specialized in diverse tasks across both 2D and 3D perception? In this paper, we define and investigate the problem of heterogeneous teacher distillation, or co-distillation, a challenging multi-teacher distillation scenario where teacher models vary significantly in both (a) their design objectives and (b) the data they were trained on. We explore data-sharing strategies and teacher-specific encoding, and introduce DUNE, a single encoder excelling in 2D vision, 3D understanding, and 3D human perception. Our model achieves performance comparable to that of its larger teachers, sometimes even outperforming them, on their respective tasks. Notably, DUNE surpasses MASt3R in Map-free Visual Relocalization with a much smaller encoder.
Authors: Bowen Tan, Ying Cui
Abstract: Fast and accurate device activity detection is the critical challenge in grant-free access for supporting massive machine-type communications (mMTC) and ultra-reliable low-latency communications (URLLC) in 5G and beyond. The state-of-the-art methods have unsatisfactory error rates or computation times. To address these outstanding issues, we propose new maximum likelihood estimation (MLE) and maximum a posterior estimation (MAPE) based device activity detection methods for known and unknown pathloss that achieve superior error rate and computation time tradeoffs using optimization and deep learning techniques. Specifically, we investigate four non-convex optimization problems for MLE and MAPE in the two pathloss cases, with one MAPE problem being formulated for the first time. For each non-convex problem, we develop an innovative parallel iterative algorithm using the parallel successive convex approximation (PSCA) method. Each PSCA-based algorithm allows parallel computations, uses up to the objective function's second-order information, converges to the problem's stationary points, and has a low per-iteration computational complexity compared to the state-of-the-art algorithms. Then, for each PSCA-based iterative algorithm, we present a deep unrolling neural network implementation, called PSCA-Net, to further reduce the computation time. Each PSCA-Net elegantly marries the underlying PSCA-based algorithm's parallel computation mechanism with the parallelizable neural network architecture and effectively optimizes its step sizes based on vast data samples to speed up the convergence. Numerical results demonstrate that the proposed methods can significantly reduce the error rate and computation time compared to the state-of-the-art methods, revealing their significant values for grant-free access.
Authors: Ron Campos, Ashmal Vayani, Parth Parag Kulkarni, Rohit Gupta, Aritra Dutta, Mubarak Shah
Abstract: Image geolocalization, in which, traditionally, an AI model predicts the precise GPS coordinates of an image is a challenging task with many downstream applications. However, the user cannot utilize the model to further their knowledge other than the GPS coordinate; the model lacks an understanding of the location and the conversational ability to communicate with the user. In recent days, with tremendous progress of large multimodal models (LMMs) -- proprietary and open-source -- researchers have attempted to geolocalize images via LMMs. However, the issues remain unaddressed; beyond general tasks, for more specialized downstream tasks, one of which is geolocalization, LMMs struggle. In this work, we propose to solve this problem by introducing a conversational model GAEA that can provide information regarding the location of an image, as required by a user. No large-scale dataset enabling the training of such a model exists. Thus we propose GAEA-1.6M, a comprehensive dataset with 800K images and around 1.6M question-answer pairs constructed by leveraging OpenStreetMap (OSM) attributes and geographical context clues. For quantitative evaluation, we propose a diverse benchmark, GAEA-Bench, comprising 4K image-text pairs to evaluate conversational capabilities equipped with diverse question types. We consider 11 state-of-the-art open-source and proprietary LMMs and demonstrate that GAEA significantly outperforms the best open-source model, LLaVA-OneVision by 25.69% and the best proprietary model, GPT-4o by 8.28%. Our dataset, model and codes are available.
Authors: Bardia Nadimi, Ghali Omar Boutaib, Hao Zheng
Abstract: Designing Verilog modules requires meticulous attention to correctness, efficiency, and adherence to design specifications. However, manually writing Verilog code remains a complex and time-consuming task that demands both expert knowledge and iterative refinement. Leveraging recent advancements in large language models (LLMs) and their structured text generation capabilities, we propose VeriMind, an agentic LLM framework for Verilog code generation that significantly automates and optimizes the synthesis process. Unlike traditional LLM-based code generators, VeriMind employs a structured reasoning approach: given a user-provided prompt describing design requirements, the system first formulates a detailed train of thought before the final Verilog code is generated. This multi-step methodology enhances interpretability, accuracy, and adaptability in hardware design. In addition, we introduce a novel evaluation metric-pass@ARC-which combines the conventional pass@k measure with Average Refinement Cycles (ARC) to capture both success rate and the efficiency of iterative refinement. Experimental results on diverse hardware design tasks demonstrated that our approach achieved up to $8.3\%$ improvement on pass@k metric and $8.1\%$ on pass@ARC metric. These findings underscore the transformative potential of agentic LLMs in automated hardware design, RTL development, and digital system synthesis.
Authors: Richie Li, Sicheng Chen
Abstract: Transformer-based large language models (LLMs) rely heavily on intensive matrix multiplications for attention and feed-forward layers, with the Q, K, and V linear projections in the Multi-Head Self-Attention (MHA) module constituting a decisive performance bottleneck. In this work, we introduce a highly optimized tiled matrix multiplication accelerator on a resource-constrained Xilinx KV260 FPGA that not only addresses this challenge but sets a new standard for efficiency and performance. Our design exploits persistent on-chip storage, a robust two-level tiling strategy for maximal data reuse, and a systolic-like unrolled compute engine that together deliver unparalleled speed and energy efficiency. Integrated with DistilBERT for Q, K, and V projections, our accelerator achieves an unequivocal 7x speedup over ARM CPU implementations (PyTorch) and an extraordinary 200x improvement over naive NumPy, reaching a throughput of up to 3.1~GFLOPs for matrix multiplications on (64,768) x (768,3072) matrices while operating at a conservative 100 MHz. These results decisively demonstrate the transformative potential of FPGA-based acceleration for critical Transformer operations, paving the way for scalable and energy-efficient deep learning inference on edge devices.
Authors: Haoming Wang, Lek-Heng Lim
Abstract: We extend the celebrated Glivenko-Cantelli theorem, sometimes called the fundamental theorem of statistics, from its standard setting of total variation distance to all $f$-divergences. A key obstacle in this endeavor is to define $f$-divergence on a subcollection of a $\sigma$-algebra that forms a $\pi$-system but not a $\sigma$-subalgebra. This is a side contribution of our work. We will show that this notion of $f$-divergence on the $\pi$-system of rays preserves nearly all known properties of standard $f$-divergence, yields a novel integral representation of the Kolmogorov-Smirnov distance, and has a Glivenko-Cantelli theorem. We will also discuss the prospects of a Vapnik-Chervonenkis theory for $f$-divergence.