Authors: Maryam Shoaeinaeini, Brent Harrison
Abstract: Human guidance in reinforcement learning (RL) is often impractical for large-scale applications due to high costs and time constraints. Large Language Models (LLMs) offer a promising alternative to mitigate RL sample inefficiency and potentially replace human trainers. However, applying LLMs as RL trainers is challenging due to their overconfidence and less reliable solutions in sequential tasks. We address this limitation by introducing a calibrated guidance system that uses Monte Carlo Dropout to enhance LLM advice reliability by assessing prediction variances from multiple forward passes. Additionally, we develop a novel RL policy shaping method based on dynamic model average entropy to adjust the LLM's influence on RL policies according to guidance uncertainty. This approach ensures robust RL training by relying on reliable LLM guidance. To validate our contributions, we conduct extensive experiments in a Minigrid environment with three goals in varying environment sizes. The results showcase superior model performance compared to uncalibrated LLMs, unguided RL, and calibrated LLMs with different shaping policies. Moreover, we analyze various uncertainty estimation methods, demonstrating the effectiveness of average entropy in reflecting higher uncertainty in incorrect guidance. These findings highlight the persistent overconfidence in fine-tuned LLMs and underscore the importance of effective calibration in sequential decision-making problems.
Authors: Pui-Wai Ma
Abstract: In our previous work [Ma and Chan (2023)], we presented a feedforward unitary equivariant neural network. We proposed three distinct activation functions tailored for this network: a softsign function with a small residue, an identity function, and a Leaky ReLU function. While these functions demonstrated the desired equivariance properties, they limited the neural network's architecture. This short paper generalises these activation functions to a single functional form. This functional form represents a broad class of functions, maintains unitary equivariance, and offers greater flexibility for the design of equivariant neural networks.
Authors: Jiang Kun
Abstract: In the previous article, we introduced a neural network framework based on symmetric differential equations. This novel framework exhibits complete symmetry, endowing it with perfect mathematical properties. While we have examined some of the system's mathematical characteristics, a detailed discussion of the network training methodology has not yet been presented. Drawing on the principles of the traditional backpropagation algorithm, this study proposes an alternative training approach that utilizes differential equation signal propagation instead of chain rule derivation. This approach not only preserves the effectiveness of training but also offers enhanced biological interpretability. The foundation of this methodology lies in the system's reversibility, which stems from its inherent symmetry,a key aspect of our research. However, this method alone is insufficient for effective neural network training. To address this, we further introduce a distributed Proportional-Integral-Derivative (PID) control approach, emphasizing its implementation within a closed system. By incorporating this method, we achieved both faster training speeds and improved accuracy. This approach not only offers novel insights into neural network training but also extends the scope of research into control methodologies. To validate its effectiveness, we apply this method to the MNIST dataset, demonstrating its practical utility.
Authors: Laurence Aitchison
Abstract: Here, we show that in the data-rich setting where you only train on each datapoint once (or equivalently, you only train for one epoch), standard "maximum likelihood" training optimizes the true data generating process (DGP) loss, which is equivalent to the test loss. Further, we show that the Bayesian model average optimizes the same objective, albeit while taking the expectation over uncertainty induced by finite data. As standard maximum likelihood training in the single-epoch setting optimizes the same objective as Bayesian inference, we argue that we do not expect Bayesian inference to offer any advantages in terms of overfitting or calibration in these settings. This explains the diminishing importance of Bayes in areas such as LLMs, which are often trained with one (or very few) epochs.
Authors: Bao Nguyen
Abstract: The thesis proposes a generalized charging framework for multiple mobile chargers to maximize the network lifetime and ensure target coverage and connectivity in large scale WRSNs. Moreover, a multi-point charging model is leveraged to enhance charging efficiency, where the MC can charge multiple sensors simultaneously at each charging location. The thesis proposes an effective Decentralized Partially Observable Semi-Markov Decision Process (Dec POSMDP) model that promotes Mobile Chargers (MCs) cooperation and detects optimal charging locations based on realtime network information. Furthermore, the proposal allows reinforcement algorithms to be applied to different networks without requiring extensive retraining. To solve the Dec POSMDP model, the thesis proposes an Asynchronous Multi Agent Reinforcement Learning algorithm (AMAPPO) based on the Proximal Policy Optimization algorithm (PPO).
Authors: Zehua Pei, Hui-Ling Zhen, Xianzhi Yu, Sinno Jialin Pan, Mingxuan Yuan, Bei Yu
Abstract: Generative Pre-trained Transformers (GPTs) have demonstrated remarkable performance across diverse domains through the extensive scaling of model parameters. Recent works observe the redundancy across the transformer blocks and develop compression methods by structured pruning of the unimportant blocks. However, such straightforward elimination will always provide irreversible performance degradation. In this paper, we propose FuseGPT, a novel methodology to recycle the pruned transformer blocks to further recover the model performance. Firstly we introduce a new importance detection metric, Macro Influence (MI), to detect the long-term influence of each transformer block by calculating their loss of information after removal. Then we propose group-level layers fusion, which adopts the parameters in layers of the unimportant blocks and injects them into the corresponding layers inside the neighboring blocks. The fusion is not one-off but through iterative parameter updates by lightweight group-level fine-tuning. Specifically, these injected parameters are frozen but weighted with learnable rank decomposition matrices to reduce the overhead during fine-tuning. Our approach not only works well on large language models but also on large multimodal models. The experiments have shown that, by using modest amounts of data, FuseGPT can outperform previous works in both perplexity and zero-shot task performance.
Authors: Aleksander Koz{\l}owski, Daniel Ponikowski, Piotr \.Zukiewicz, Pawe{\l} Twardowski
Abstract: We propose an End-to-end Convolutional Activation Anomaly Analysis (E2E-CA$^3$), which is a significant extension of A$^3$ anomaly detection approach proposed by Sperl, Schulze and B\"ottinger, both in terms of architecture and scope of application. In contrast to the original idea, we utilize a convolutional autoencoder as a target network, which allows for natural application of the method both to image and tabular data. The alarm network is also designed as a CNN, where the activations of convolutional layers from CAE are stacked together into $k+1-$dimensional tensor. Moreover, we combine the classification loss of the alarm network with the reconstruction error of the target CAE, as a "best of both worlds" approach, which greatly increases the versatility of the network. The evaluation shows that despite generally straightforward and lightweight architecture, it has a very promising anomaly detection performance on common datasets such as MNIST, CIFAR-10 and KDDcup99.
Authors: Mayank Nautiyal, Andrey Shternshis, Andreas Hellander, Prashant Singh
Abstract: We present a generative modeling approach based on the variational inference framework for likelihood-free simulation-based inference. The method leverages latent variables within variational autoencoders to efficiently estimate complex posterior distributions arising from stochastic simulations. We explore two variations of this approach distinguished by their treatment of the prior distribution. The first model adapts the prior based on observed data using a multivariate prior network, enhancing generalization across various posterior queries. In contrast, the second model utilizes a standard Gaussian prior, offering simplicity while still effectively capturing complex posterior distributions. We demonstrate the efficacy of these models on well-established benchmark problems, achieving results comparable to flow-based approaches while maintaining computational efficiency and scalability.
Authors: Tri Cao, Minh-Huy Trinh, Ailin Deng, Quoc-Nam Nguyen, Khoa Duong, Ngai-Man Cheung, Bryan Hooi
Abstract: Anomaly detection (AD) is a machine learning task that identifies anomalies by learning patterns from normal training data. In many real-world scenarios, anomalies vary in severity, from minor anomalies with little risk to severe abnormalities requiring immediate attention. However, existing models primarily operate in a binary setting, and the anomaly scores they produce are usually based on the deviation of data points from normal data, which may not accurately reflect practical severity. In this paper, we address this gap by making three key contributions. First, we propose a novel setting, Multilevel AD (MAD), in which the anomaly score represents the severity of anomalies in real-world applications, and we highlight its diverse applications across various domains. Second, we introduce a novel benchmark, MAD-Bench, that evaluates models not only on their ability to detect anomalies, but also on how effectively their anomaly scores reflect severity. This benchmark incorporates multiple types of baselines and real-world applications involving severity. Finally, we conduct a comprehensive performance analysis on MAD-Bench. We evaluate models on their ability to assign severity-aligned scores, investigate the correspondence between their performance on binary and multilevel detection, and study their robustness. This analysis offers key insights into improving AD models for practical severity alignment. The code framework and datasets used for the benchmark will be made publicly available.
Authors: Shenglai Zeng, Jiankun Zhang, Bingheng Li, Yuping Lin, Tianqi Zheng, Dante Everaert, Hanqing Lu, Hui Liu, Hui Liu, Yue Xing, Monica Xiao Cheng, Jiliang Tang
Abstract: Retrieval-Augmented Generation (RAG) systems have shown promise in enhancing the performance of Large Language Models (LLMs). However, these systems face challenges in effectively integrating external knowledge with the LLM's internal knowledge, often leading to issues with misleading or unhelpful information. This work aims to provide a systematic study on knowledge checking in RAG systems. We conduct a comprehensive analysis of LLM representation behaviors and demonstrate the significance of using representations in knowledge checking. Motivated by the findings, we further develop representation-based classifiers for knowledge filtering. We show substantial improvements in RAG performance, even when dealing with noisy knowledge databases. Our study provides new insights into leveraging LLM representations for enhancing the reliability and effectiveness of RAG systems.
Authors: Sanaz Mahmoodi Takaghaj, Jack Sampson
Abstract: Recent advancements in machine learning, particularly through deep learning architectures like PointNet, have transformed the processing of three-dimensional (3D) point clouds, significantly improving 3D object classification and segmentation tasks. While 3D point clouds provide detailed spatial information, spatio-temporal signals introduce a dynamic element that accounts for changes over time. However, applying deep learning techniques to spatio-temporal signals and deploying them on edge devices presents challenges, including real-time processing, memory capacity, and power consumption. To address these issues, this paper presents a novel approach that combines PointNet's feature extraction with the in-memory computing capabilities and energy efficiency of neuromorphic systems for spatio-temporal signal recognition. The proposed method consists of a two-stage process: in the first stage, PointNet extracts features from the spatio-temporal signals, which are then stored in non-volatile memristor crossbar arrays. In the second stage, these features are processed by a single-layer spiking neural encoder-decoder that employs the Locally Competitive Algorithm (LCA) for efficient encoding and classification. This work integrates the strengths of both PointNet and LCA, enhancing computational efficiency and energy performance on edge devices. PointLCA-Net achieves high recognition accuracy for spatio-temporal data with substantially lower energy burden during both inference and training than comparable approaches, thus advancing the deployment of advanced neural architectures in energy-constrained environments.
Authors: SungHeon Jeong, Hamza Errahmouni Barkam, Sanggeon Yun, Yeseong Kim, Shaahin Angizi, Mohsen Imani
Abstract: Hyperdimensional computing (HDC) enables efficient data encoding and processing in high-dimensional space, benefiting machine learning and data analysis. However, underutilization of these spaces can lead to overfitting and reduced model reliability, especially in data-limited systems a critical issue in sectors like healthcare that demand robustness and consistent performance. We introduce BoostHD, an approach that applies boosting algorithms to partition the hyperdimensional space into subspaces, creating an ensemble of weak learners. By integrating boosting with HDC, BoostHD enhances performance and reliability beyond existing HDC methods. Our analysis highlights the importance of efficient utilization of hyperdimensional spaces for improved model performance. Experiments on healthcare datasets show that BoostHD outperforms state-of-the-art methods. On the WESAD dataset, it achieved an accuracy of 98.37%, surpassing Random Forest, XGBoost, and OnlineHD. BoostHD also demonstrated superior inference efficiency and stability, maintaining high accuracy under data imbalance and noise. In person-specific evaluations, it achieved an average accuracy of 96.19%, outperforming other models. By addressing the limitations of both boosting and HDC, BoostHD expands the applicability of HDC in critical domains where reliability and precision are paramount.
Authors: Vincent Mai, Quang Hung Pham, Arthur Favrel, Jean-Philippe Gauthier, Martin Gagnon
Abstract: Hydro-generating units (HGUs) play a crucial role in integrating intermittent renewable energy sources into the power grid due to their flexible operational capabilities. This evolving role has led to an increase in transient events, such as startups, which impose significant stresses on turbines, leading to increased turbine fatigue and a reduced operational lifespan. Consequently, optimizing startup sequences to minimize stresses is vital for hydropower utilities. However, this task is challenging, as stress measurements on prototypes can be expensive and time-consuming. To tackle this challenge, we propose an innovative automated approach to optimize the startup parameters of HGUs with a limited budget of measured startup sequences. Our method combines active learning and black-box optimization techniques, utilizing virtual strain sensors and dynamic simulations of HGUs. This approach was tested in real-time during an on-site measurement campaign on an instrumented Francis turbine prototype. The results demonstrate that our algorithm successfully identified an optimal startup sequence using only seven measured sequences. It achieves a remarkable 42% reduction in the maximum strain cycle amplitude compared to the standard startup sequence. This study paves the way for more efficient HGU startup optimization, potentially extending their operational lifespans.
Authors: Demian Pavlyshenko, Bohdan Pavlyshenko
Abstract: The paper considers exploratory data analysis and approaches in predictive analytics for air alerts during the Russian-Ukrainian war which broke out on Feb 24, 2022. The results illustrate that alerts in regions correlate with one another and have geospatial patterns which make it feasible to build a predictive model which predicts alerts that are expected to take place in a certain region within a specified time period. The obtained results show that the alert status in a particular region is highly dependable on the features of its adjacent regions. Seasonality features like hours, days of a week and months are also crucial in predicting the target variable. Some regions highly rely on the time feature which equals to a number of days from the initial date of the dataset. From this, we can deduce that the air alert pattern changes throughout the time.
Authors: Armani Rodriguez, Silvija Kokalj-Filipovic
Abstract: Generating high-quality speech efficiently remains a key challenge for generative models in speech synthesis. This paper introduces VQalAttent, a lightweight model designed to generate fake speech with tunable performance and interpretability. Leveraging the AudioMNIST dataset, consisting of human utterances of decimal digits (0-9), our method employs a two-step architecture: first, a scalable vector quantized autoencoder (VQ-VAE) that compresses audio spectrograms into discrete latent representations, and second, a decoder-only transformer that learns the probability model of these latents. Trained transformer generates similar latent sequences, convertible to audio spectrograms by the VQ-VAE decoder, from which we generate fake utterances. Interpreting statistical and perceptual quality of the fakes, depending on the dimension and the extrinsic information of the latent space, enables guided improvements in larger, commercial generative models. As a valuable tool for understanding and refining audio synthesis, our results demonstrate VQalAttent's capacity to generate intelligible speech samples with limited computational resources, while the modularity and transparency of the training pipeline helps easily correlate the analytics with modular modifications, hence providing insights for the more complex models.
Authors: Minghua Wang, Ziyun Huang, Jinhui Xu
Abstract: To improve persistence diagram representation learning, we propose Multiset Transformer. This is the first neural network that utilizes attention mechanisms specifically designed for multisets as inputs and offers rigorous theoretical guarantees of permutation invariance. The architecture integrates multiset-enhanced attentions with a pool-decomposition scheme, allowing multiplicities to be preserved across equivariant layers. This capability enables full leverage of multiplicities while significantly reducing both computational and spatial complexity compared to the Set Transformer. Additionally, our method can greatly benefit from clustering as a preprocessing step to further minimize complexity, an advantage not possessed by the Set Transformer. Experimental results demonstrate that the Multiset Transformer outperforms existing neural network methods in the realm of persistence diagram representation learning.
Authors: Tengjie Zheng, Lin Cheng, Shengping Gong, Xu Huang
Abstract: Learning dynamical models from data is not only fundamental but also holds great promise for advancing principle discovery, time-series prediction, and controller design. Among various approaches, Gaussian Process State-Space Models (GPSSMs) have recently gained significant attention due to their combination of flexibility and interpretability. However, for online learning, the field lacks an efficient method suitable for scenarios where prior information regarding data distribution and model function is limited. To address this issue, this paper proposes a recursive GPSSM method with adaptive capabilities for both operating domains and Gaussian process (GP) hyperparameters. Specifically, we first utilize first-order linearization to derive a Bayesian update equation for the joint distribution between the system state and the GP model, enabling closed-form and domain-independent learning. Second, an online selection algorithm for inducing points is developed based on informative criteria to achieve lightweight learning. Third, to support online hyperparameter optimization, we recover historical measurement information from the current filtering distribution. Comprehensive evaluations on both synthetic and real-world datasets demonstrate the superior accuracy, computational efficiency, and adaptability of our method compared to state-of-the-art online GPSSM techniques.
Authors: Matthew Spellings, Maya Martirossyan, Julia Dshemuchadse
Abstract: Recent work has proven that training large language models with self-supervised tasks and fine-tuning these models to complete new tasks in a transfer learning setting is a powerful idea, enabling the creation of models with many parameters, even with little labeled data; however, the number of domains that have harnessed these advancements has been limited. In this work, we formulate a set of geometric tasks suitable for the large-scale study of ordered three-dimensional structures, without requiring any human intervention in data labeling. We build deep rotation- and permutation-equivariant neural networks based on geometric algebra and use them to solve these tasks on both idealized and simulated three-dimensional structures. Quantifying order in complex-structured assemblies remains a long-standing challenge in materials physics; these models can elucidate the behavior of real self-assembling systems in a variety of ways, from distilling insights from learned tasks without further modification to solving new tasks with smaller amounts of labeled data via transfer learning.
Authors: Hansol Lim, Jee Won Lee, Jonathan Boyack, Jongseong Brad Choi
Abstract: An onboard prediction of dynamic parameters (e.g. Aerodynamic drag, rolling resistance) enables accurate path planning for EVs. This paper presents EV-PINN, a Physics-Informed Neural Network approach in predicting instantaneous battery power and cumulative energy consumption during cruising while generalizing to the nonlinear dynamics of an EV. Our method learns real-world parameters such as motor efficiency, regenerative braking efficiency, vehicle mass, coefficient of aerodynamic drag, and coefficient of rolling resistance using automatic differentiation based on dynamics and ensures consistency with ground truth vehicle data. EV-PINN was validated using 15 and 35 minutes of in-situ battery log data from the Tesla Model 3 Long Range and Tesla Model S, respectively. With only vehicle speed and time as inputs, our model achieves high accuracy and generalization to dynamics, with validation losses of 0.002195 and 0.002292, respectively. This demonstrates EV-PINN's effectiveness in estimating parameters and predicting battery usage under actual driving conditions without the need for additional sensors.
Authors: Eric Tang, Bangding Yang, Xingyou Song
Abstract: With the rise of large language models (LLMs) for flexibly processing information as strings, a natural application is regression, specifically by preprocessing string representations into LLM embeddings as downstream features for metric prediction. In this paper, we provide one of the first comprehensive investigations into embedding-based regression and demonstrate that LLM embeddings as features can be better for high-dimensional regression tasks than using traditional feature engineering. This regression performance can be explained in part due to LLM embeddings over numeric data inherently preserving Lipschitz continuity over the feature space. Furthermore, we quantify the contribution of different model effects, most notably model size and language understanding, which we find surprisingly do not always improve regression performance.
Authors: Binqian Xu, Xiangbo Shu, Haiyang Mei, Guosen Xie, Basura Fernando, Mike Zheng Shou, Jinhui Tang
Abstract: Multimodal Large Language Models (MLLMs) have made significant advancements, demonstrating powerful capabilities in processing and understanding multimodal data. Fine-tuning MLLMs with Federated Learning (FL) allows for expanding the training data scope by including private data sources, thereby enhancing their practical applicability in privacy-sensitive domains. However, current research remains in the early stage, particularly in addressing the \textbf{multimodal heterogeneities} in real-world applications. In this paper, we introduce a benchmark for evaluating various downstream tasks in the federated fine-tuning of MLLMs within multimodal heterogeneous scenarios, laying the groundwork for the research in the field. Our benchmark encompasses two datasets, five comparison baselines, and four multimodal scenarios, incorporating over ten types of modal heterogeneities. To address the challenges posed by modal heterogeneity, we develop a general FedMLLM framework that integrates four representative FL methods alongside two modality-agnostic strategies. Extensive experimental results show that our proposed FL paradigm improves the performance of MLLMs by broadening the range of training data and mitigating multimodal heterogeneity. Code is available at https://github.com/1xbq1/FedMLLM
Authors: Xiangyu Zhang
Abstract: The generation of drug-like molecules is crucial for drug design. Existing reinforcement learning (RL) methods often overlook structural information. However, feature engineering-based methods usually merely focus on binding affinity prediction without substantial molecular modification. To address this, we present Graph-based Topological Reinforcement Learning (GraphTRL), which integrates both chemical and structural data for improved molecular generation. GraphTRL leverages multiscale weighted colored graphs (MWCG) and persistent homology, combined with molecular fingerprints, as the state space for RL. Evaluations show that GraphTRL outperforms existing methods in binding affinity prediction, offering a promising approach to accelerate drug discovery.
Authors: Junyang Chen, Yiqun Zhang, Mengke Li, Yang Lu, Yiu-ming Cheung
Abstract: Clustering complex data in the form of attributed graphs has attracted increasing attention, where appropriate graph representation is a critical prerequisite for accurate cluster analysis. However, the Graph Convolutional Network will homogenize the representation of graph nodes due to the well-known over-smoothing effect. This limits the network architecture to a shallow one, losing the ability to capture the critical global distribution information for clustering. Therefore, we propose a generalized graph auto-encoder network, which introduces quaternion operations to the encoders to achieve efficient structured feature representation learning without incurring deeper network and larger-scale parameters. The generalization of our method lies in the following two aspects: 1) connecting the quaternion operation naturally suitable for four feature components with graph data of arbitrary attribute dimensions, and 2) introducing a generalized graph clustering objective as a loss term to obtain clustering-friendly representations without requiring a pre-specified number of clusters $k$. It turns out that the representations of nodes learned by the proposed Graph Clustering based on Generalized Quaternion representation learning (GCGQ) are more discriminative, containing global distribution information, and are more general, suiting downstream clustering under different $k$s. Extensive experiments including significance tests, ablation studies, and qualitative results, illustrate the superiority of GCGQ. The source code is temporarily opened at \url{https://anonymous.4open.science/r/ICLR-25-No7181-codes}.
URLs: https://anonymous.4open.science/r/ICLR-25-No7181-codes
Authors: Gabriel Santos, Rita Julia, Marcelo Nascimento
Abstract: Clustering data using prior domain knowledge, starting from a partially labeled set, has recently been widely investigated. Often referred to as semi-supervised clustering, this approach leverages labeled data to enhance clustering accuracy. To maximize algorithm performance, it is crucial to ensure the safety of this prior knowledge. Methods addressing this concern are termed safe semi-supervised clustering (S3C) algorithms. This paper introduces the KNN graph-based safety-aware semi-supervised fuzzy c-means algorithm (K-GBS3FCM), which dynamically assesses neighborhood relationships between labeled and unlabeled data using the K-Nearest Neighbors (KNN) algorithm. This approach aims to optimize the use of labeled data while minimizing the adverse effects of incorrect labels. Additionally, it is proposed a mechanism that adjusts the influence of labeled data on unlabeled ones through regularization parameters and the average safety degree. Experimental results on multiple benchmark datasets demonstrate that the graph-based approach effectively leverages prior knowledge to enhance clustering accuracy. The proposed method was significantly superior in 64% of the 56 test configurations, obtaining higher levels of clustering accuracy when compared to other semi-supervised and traditional unsupervised methods. This research highlights the potential of integrating graph-based approaches, such as KNN, with established techniques to develop advanced clustering algorithms, offering significant applications in fields that rely on both labeled and unlabeled data for more effective clustering.
Authors: Donghyeon Yi, Seoyoung Lee, Jongho Kim, Junyoung Kim, Sohmyung Ha, Ik Joon Chang, Minkyu Je
Abstract: Encoder-based transformers, powered by self-attention layers, have revolutionized machine learning with their context-aware representations. However, their quadratic growth in computational and memory demands presents significant bottlenecks. Analog-Mixed-Signal Process-in-Memory (AMS-PiM) architectures address these challenges by enabling efficient on-chip processing. Traditionally, AMS-PiM relies on Quantization-Aware Training (QAT), which is hardware-efficient but requires extensive retraining to adapt models to AMS-PiMs, making it increasingly impractical for transformer models. Post-Training Quantization (PTQ) mitigates this training overhead but introduces significant hardware inefficiencies. PTQ relies on dequantization-quantization (DQ-Q) processes, floating-point units (FPUs), and high-ENOB (Effective Number of Bits) analog-to-digital converters (ADCs). Particularly, High-ENOB ADCs scale exponentially in area and energy ($2^{ENOB}$), reduce sensing margins, and increase susceptibility to process, voltage, and temperature (PVT) variations, further compounding PTQ's challenges in AMS-PiM systems. To overcome these limitations, we propose RAP, an AMS-PiM architecture that eliminates DQ-Q processes, introduces FPU- and division-free nonlinear processing, and employs a low-ENOB-ADC-based sparse Matrix Vector multiplication technique. Using the proposed techniques, RAP improves error resiliency, area/energy efficiency, and computational speed while preserving numerical stability. Experimental results demonstrate that RAP outperforms state-of-the-art GPUs and conventional PiM architectures in energy efficiency, latency, and accuracy, making it a scalable solution for the efficient deployment of transformers.
Authors: Kai Lu, Siqi Zhao, Jiguang Wan
Abstract: Efficient management of storage resources in big data and cloud computing environments requires accurate identification of data's "cold" and "hot" states. Traditional methods, such as rule-based algorithms and early AI techniques, often struggle with dynamic workloads, leading to low accuracy, poor adaptability, and high operational overhead. To address these issues, we propose a novel solution based on online learning strategies. Our approach dynamically adapts to changing data access patterns, achieving higher accuracy and lower operational costs. Rigorous testing with both synthetic and real-world datasets demonstrates a significant improvement, achieving a 90% accuracy rate in hot-cold classification. Additionally, the computational and storage overheads are considerably reduced.
Authors: Stefan K. Nielsen, Tan M. Nguyen
Abstract: Contrastive learning has proven instrumental in learning unbiased representations of data, especially in complex environments characterized by high-cardinality and high-dimensional sensitive information. However, existing approaches within this setting require predefined modelling assumptions of bias-causing interactions that limit the model's ability to learn debiased representations. In this work, we propose a new method for fair contrastive learning that employs an attention mechanism to model bias-causing interactions, enabling the learning of a fairer and semantically richer embedding space. In particular, our attention mechanism avoids bias-causing samples that confound the model and focuses on bias-reducing samples that help learn semantically meaningful representations. We verify the advantages of our method against existing baselines in fair contrastive learning and show that our approach can significantly boost bias removal from learned representations without compromising downstream accuracy.
Authors: Silin Zhou, Shuo Shang, Lisi Chen, Peng Han, Christian S. Jensen
Abstract: Trajectory representation learning (TRL) maps trajectories to vectors that can be used for many downstream tasks. Existing TRL methods use either grid trajectories, capturing movement in free space, or road trajectories, capturing movement in a road network, as input. We observe that the two types of trajectories are complementary, providing either region and location information or providing road structure and movement regularity. Therefore, we propose a novel multimodal TRL method, dubbed GREEN, to jointly utilize Grid and Road trajectory Expressions for Effective representatioN learning. In particular, we transform raw GPS trajectories into both grid and road trajectories and tailor two encoders to capture their respective information. To align the two encoders such that they complement each other, we adopt a contrastive loss to encourage them to produce similar embeddings for the same raw trajectory and design a mask language model (MLM) loss to use grid trajectories to help reconstruct masked road trajectories. To learn the final trajectory representation, a dual-modal interactor is used to fuse the outputs of the two encoders via cross-attention. We compare GREEN with 7 state-of-the-art TRL methods for 3 downstream tasks, finding that GREEN consistently outperforms all baselines and improves the accuracy of the best-performing baseline by an average of 15.99\%.
Authors: Mahammad Humayoo
Abstract: In numerous episodic reinforcement learning (RL) settings, SARSA-based methodologies are employed to enhance policies aimed at maximizing returns over long horizons. Conventional SARSA algorithms, however, have difficulties in balancing bias and variation due to the reliance on a singular, fixed discount factor. This study expands the temporal difference decomposition approach, TD($\triangle$), to the SARSA algorithm. SARSA, a widely utilised on-policy RL method, enhances action-value functions via temporal difference updates. TD($\triangle$) facilitates learning over several time-scales by breaking the action-value function into components associated with distinct discount factors. This decomposition improves learning efficiency and stability, particularly in problems necessitating long-horizon optimization. We illustrate that our methodology mitigates bias in SARSA's updates while facilitating accelerated convergence in contexts characterized by dense rewards. Experimental findings across many benchmark tasks indicate that the proposed SARSA($\triangle$) surpasses conventional TD learning methods in both tabular and deep RL contexts.
Authors: Hongbo Liu
Abstract: Contrastive Language-Image Pre-training (CLIP) has attracted a surge of attention for its superior zero-shot performance and excellent transferability to downstream tasks. However, training such large-scale models usually requires substantial computation and storage, which poses barriers for general users with consumer-level computers. Motivated by this observation, in this paper we investigate how to achieve competitive performance on only one Nvidia RTX3090 GPU and with one terabyte for storing dataset. On one hand, we simplify the transformer block structure and combine Weight Inheritance with multi-stage Knowledge Distillation (WIKD), thereby reducing the parameters and improving the inference speed during training along with deployment. On the other hand, confronted with the convergence challenge posed by small dataset, we generate synthetic captions for each sample as data augmentation, and devise a novel Pair Matching (PM) loss to fully exploit the distinguishment among positive and negative image-text pairs. Extensive experiments demonstrate that our model can achieve a new state-of-the-art datascale-parameter-accuracy tradeoff, which could further popularize the CLIP model in the related research community.
Authors: Ke Zhu, Yu Wang, Yanpeng Sun, Qiang Chen, Jiangjiang Liu, Gang Zhang, Jingdong Wang
Abstract: Multimodal RLHF usually happens after supervised finetuning (SFT) stage to continually improve vision-language models' (VLMs) comprehension. Conventional wisdom holds its superiority over continual SFT during this preference alignment stage. In this paper, we observe that the inherent value of multimodal RLHF lies in its negative supervision, the logit of the rejected responses. We thus propose a novel negative supervised finetuning (nSFT) approach that fully excavates these information resided. Our nSFT disentangles this negative supervision in RLHF paradigm, and continually aligns VLMs with a simple SFT loss. This is more memory efficient than multimodal RLHF where 2 (e.g., DPO) or 4 (e.g., PPO) large VLMs are strictly required. The effectiveness of nSFT is rigorously proved by comparing it with various multimodal RLHF approaches, across different dataset sources, base VLMs and evaluation metrics. Besides, fruitful of ablations are provided to support our hypothesis. We hope this paper will stimulate further research to properly align large vision language models.
Authors: Jie Zhang, Kristina Nikoli\'c, Nicholas Carlini, Florian Tram\`er
Abstract: Ensemble everything everywhere is a defense to adversarial examples that was recently proposed to make image classifiers robust. This defense works by ensembling a model's intermediate representations at multiple noisy image resolutions, producing a single robust classification. This defense was shown to be effective against multiple state-of-the-art attacks. Perhaps even more convincingly, it was shown that the model's gradients are perceptually aligned: attacks against the model produce noise that perceptually resembles the targeted class. In this short note, we show that this defense is not robust to adversarial attack. We first show that the defense's randomness and ensembling method cause severe gradient masking. We then use standard adaptive attack techniques to reduce the defense's robust accuracy from 48% to 1% on CIFAR-100 and from 62% to 4% on CIFAR-10, under the $\ell_\infty$-norm threat model with $\varepsilon=8/255$.
Authors: Teodor Alexandru Szente, James Harrison, Mihai Zanfir, Cristian Sminchisescu
Abstract: Fractional gradient descent has been studied extensively, with a focus on its ability to extend traditional gradient descent methods by incorporating fractional-order derivatives. This approach allows for more flexibility in navigating complex optimization landscapes and offers advantages in certain types of problems, particularly those involving non-linearities and chaotic dynamics. Yet, the challenge of fine-tuning the fractional order parameters remains unsolved. In this work, we demonstrate that it is possible to train a neural network to predict the order of the gradient effectively.
Authors: Giung Nam, Juho Lee
Abstract: While ensembling deep neural networks has shown promise in improving generalization performance, scaling current ensemble methods for large models remains challenging. Given that recent progress in deep learning is largely driven by the scale, exemplified by the widespread adoption of large-scale neural network architectures, scalability emerges an increasingly critical issue for machine learning algorithms in the era of large-scale models. In this work, we first showcase the potential of low precision ensembling, where ensemble members are derived from a single model within low precision number systems in a training-free manner. Our empirical analysis demonstrates the effectiveness of our proposed low precision ensembling method compared to existing ensemble approaches.
Authors: Irina Bara\v{s}in, Bla\v{z} Bertalani\v{c}, Miha Mohor\v{c}i\v{c}, Carolina Fortuna
Abstract: Time series classification is a relevant step supporting decision-making processes in various domains, and deep neural models have shown promising performance. Despite significant advancements in deep learning, the theoretical understanding of how and why complex architectures function remains limited, prompting the need for more interpretable models. Recently, the Kolmogorov-Arnold Networks (KANs) have been proposed as a more interpretable alternative. While KAN-related research is significantly rising, to date, the study of KAN architectures for time series classification has been limited. In this paper, we aim to conduct a comprehensive and robust exploration of the KAN architecture for time series classification on the UCR benchmark. More specifically, we look at a) how reference architectures for forecasting transfer to classification, at the b) hyperparameter and implementation influence on the classification performance in view of finding the one that performs best on the selected benchmark, the c) complexity trade-offs and d) interpretability advantages. Our results show that (1) Efficient KAN outperforms MLP in performance and computational efficiency, showcasing its suitability for tasks classification tasks. (2) Efficient KAN is more stable than KAN across grid sizes, depths, and layer configurations, particularly with lower learning rates. (3) KAN maintains competitive accuracy compared to state-of-the-art models like HIVE-COTE2, with smaller architectures and faster training times, supporting its balance of performance and transparency. (4) The interpretability of the KAN model aligns with findings from SHAP analysis, reinforcing its capacity for transparent decision-making.
Authors: Mohammed Akkaoui
Abstract: Muscular Dystrophy is a group of genetic disorders that progressively affect the strength and functioning of muscles, thereby affecting millions of people worldwide. The lifetime nature of MD requires continuous follow-up care due to its progressive nature. This conceptual paper proposes an Internet of Things-based system to support the management of MD through remote, multi-dimensional monitoring of patients in order to provide real-time health status updates. Traditional methods have failed to give actionable data in real time, hence denying healthcare providers the opportunity to make evidence-based decisions. Technology-driven approaches are urgently needed to provide deep insights into disease progression and patient health. It aims to enhance treatment strategies, enabling patients to better manage their condition and giving healthcare professionals more confidence in their management decisions.
Authors: Junjie Shan, Ziqi Zhao, Jialin Lu, Rui Zhang, Siu Ming Yiu, Ka-Ho Chow
Abstract: Foundation models that bridge vision and language have made significant progress, inspiring numerous life-enriching applications. However, their potential for misuse to introduce new threats remains largely unexplored. This paper reveals that vision-language models (VLMs) can be exploited to overcome longstanding limitations in gradient inversion attacks (GIAs) within federated learning (FL), where an FL server reconstructs private data samples from gradients shared by victim clients. Current GIAs face challenges in reconstructing high-resolution images, especially when the victim has a large local data batch. While focusing reconstruction on valuable samples rather than the entire batch is promising, existing methods lack the flexibility to allow attackers to specify their target data. In this paper, we introduce Geminio, the first approach to transform GIAs into semantically meaningful, targeted attacks. Geminio enables a brand new privacy attack experience: attackers can describe, in natural language, the types of data they consider valuable, and Geminio will prioritize reconstruction to focus on those high-value samples. This is achieved by leveraging a pretrained VLM to guide the optimization of a malicious global model that, when shared with and optimized by a victim, retains only gradients of samples that match the attacker-specified query. Extensive experiments demonstrate Geminio's effectiveness in pinpointing and reconstructing targeted samples, with high success rates across complex datasets under FL and large batch sizes and showing resilience against existing defenses.
Authors: Joseph Farrington, Samah Alimam, Martin Utley, Kezhi Li, Wai Keong Wong
Abstract: Efforts to reduce platelet wastage in hospital blood banks have focused on ordering policies, but the predominant practice of issuing the oldest unit first may not be optimal when some units are returned unused. We propose a novel, machine learning (ML)-guided issuing policy to increase the likelihood of returned units being reissued before expiration. Our ML model trained to predict returns on 17,297 requests for platelets gave AUROC 0.74 on 9,353 held-out requests. Prior to ML model development we built a simulation of the blood bank operation that incorporated returns to understand the scale of benefits of such a model. Using our trained model in the simulation gave an estimated reduction in wastage of 14%. Our partner hospital is considering adopting our approach, which would be particularly beneficial for hospitals with higher return rates and where units have a shorter remaining useful life on arrival.
Authors: Jieming Bian, Lei Wang, Letian Zhang, Jie Xu
Abstract: Foundation models (FMs) achieve strong performance across diverse tasks with task-specific fine-tuning, yet full parameter fine-tuning is often computationally prohibitive for large models. Parameter-efficient fine-tuning (PEFT) methods like Low-Rank Adaptation (LoRA) reduce this cost by introducing low-rank matrices for tuning fewer parameters. While LoRA allows for efficient fine-tuning, it requires significant data for adaptation, making Federated Learning (FL) an appealing solution due to its privacy-preserving collaborative framework. However, combining LoRA with FL introduces two key challenges: the \textbf{Server-Side LoRA Aggregation Bias}, where server-side averaging of LoRA matrices diverges from the ideal global update, and the \textbf{Client-Side LoRA Initialization Drift}, emphasizing the need for consistent initialization across rounds. Existing approaches address these challenges individually, limiting their effectiveness. We propose LoRA-FAIR, a novel method that tackles both issues by introducing a correction term on the server while keeping the original LoRA modules, enhancing aggregation efficiency and accuracy. LoRA-FAIR maintains computational and communication efficiency, yielding superior performance over state-of-the-art methods. Experimental results on ViT and MLP-Mixer models across large-scale datasets demonstrate that LoRA-FAIR consistently achieves performance improvements in FL settings.
Authors: Colin Diggs, Michael Doyle, Amit Madan, Siggy Scott, Emily Escamilla, Jacob Zimmer, Naveed Nekoo, Paul Ursino, Michael Bartholf, Zachary Robin, Anand Patel, Chris Glasz, William Macke, Paul Kirk, Jasper Phillips, Arun Sridharan, Doug Wendt, Scott Rosen, Nitin Naik, Justin F. Brunelle, Samruddhi Thaker
Abstract: Legacy software systems, written in outdated languages like MUMPS and mainframe assembly, pose challenges in efficiency, maintenance, staffing, and security. While LLMs offer promise for modernizing these systems, their ability to understand legacy languages is largely unknown. This paper investigates the utilization of LLMs to generate documentation for legacy code using two datasets: an electronic health records (EHR) system in MUMPS and open-source applications in IBM mainframe Assembly Language Code (ALC). We propose a prompting strategy for generating line-wise code comments and a rubric to evaluate their completeness, readability, usefulness, and hallucination. Our study assesses the correlation between human evaluations and automated metrics, such as code complexity and reference-based metrics. We find that LLM-generated comments for MUMPS and ALC are generally hallucination-free, complete, readable, and useful compared to ground-truth comments, though ALC poses challenges. However, no automated metrics strongly correlate with comment quality to predict or measure LLM performance. Our findings highlight the limitations of current automated measures and the need for better evaluation metrics for LLM-generated documentation in legacy systems.
Authors: Patrik Kenfack, Ulrich A\"ivodji, Samira Ebrahimi Kahou
Abstract: Neural networks can learn spurious correlations in the data, often leading to performance disparity for underrepresented subgroups. Studies have demonstrated that the disparity is amplified when knowledge is distilled from a complex teacher model to a relatively "simple" student model. Prior work has shown that ensemble deep learning methods can improve the performance of the worst-case subgroups; however, it is unclear if this advantage carries over when distilling knowledge from an ensemble of teachers, especially when the teacher models are debiased. This study demonstrates that traditional ensemble knowledge distillation can significantly drop the performance of the worst-case subgroups in the distilled student model even when the teacher models are debiased. To overcome this, we propose Adaptive Group Robust Ensemble Knowledge Distillation (AGRE-KD), a simple ensembling strategy to ensure that the student model receives knowledge beneficial for unknown underrepresented subgroups. Leveraging an additional biased model, our method selectively chooses teachers whose knowledge would better improve the worst-performing subgroups by upweighting the teachers with gradient directions deviating from the biased model. Our experiments on several datasets demonstrate the superiority of the proposed ensemble distillation technique and show that it can even outperform classic model ensembles based on majority voting.
Authors: Guojun Xiong, Shufan Wang, Daniel Jiang, Jian Li
Abstract: Federated reinforcement learning (FedRL) enables multiple agents to collaboratively learn a policy without sharing their local trajectories collected during agent-environment interactions. However, in practice, the environments faced by different agents are often heterogeneous, leading to poor performance by the single policy learned by existing FedRL algorithms on individual agents. In this paper, we take a further step and introduce a \emph{personalized} FedRL framework (PFedRL) by taking advantage of possibly shared common structure among agents in heterogeneous environments. Specifically, we develop a class of PFedRL algorithms named PFedRL-Rep that learns (1) a shared feature representation collaboratively among all agents, and (2) an agent-specific weight vector personalized to its local environment. We analyze the convergence of PFedTD-Rep, a particular instance of the framework with temporal difference (TD) learning and linear representations. To the best of our knowledge, we are the first to prove a linear convergence speedup with respect to the number of agents in the PFedRL setting. To achieve this, we show that PFedTD-Rep is an example of the federated two-timescale stochastic approximation with Markovian noise. Experimental results demonstrate that PFedTD-Rep, along with an extension to the control setting based on deep Q-networks (DQN), not only improve learning in heterogeneous settings, but also provide better generalization to new environments.
Authors: Zeyang Li, Navid Azizan
Abstract: Multi-agent reinforcement learning (MARL) has achieved notable success in cooperative tasks, demonstrating impressive performance and scalability. However, deploying MARL agents in real-world applications presents critical safety challenges. Current safe MARL algorithms are largely based on the constrained Markov decision process (CMDP) framework, which enforces constraints only on discounted cumulative costs and lacks an all-time safety assurance. Moreover, these methods often overlook the feasibility issue (the system will inevitably violate state constraints within certain regions of the constraint set), resulting in either suboptimal performance or increased constraint violations. To address these challenges, we propose a novel theoretical framework for safe MARL with $\textit{state-wise}$ constraints, where safety requirements are enforced at every state the agents visit. To resolve the feasibility issue, we leverage a control-theoretic notion of the feasible region, the controlled invariant set (CIS), characterized by the safety value function. We develop a multi-agent method for identifying CISs, ensuring convergence to a Nash equilibrium on the safety value function. By incorporating CIS identification into the learning process, we introduce a multi-agent dual policy iteration algorithm that guarantees convergence to a generalized Nash equilibrium in state-wise constrained cooperative Markov games, achieving an optimal balance between feasibility and performance. Furthermore, for practical deployment in complex high-dimensional systems, we propose $\textit{Multi-Agent Dual Actor-Critic}$ (MADAC), a safe MARL algorithm that approximates the proposed iteration scheme within the deep RL paradigm. Empirical evaluations on safe MARL benchmarks demonstrate that MADAC consistently outperforms existing methods, delivering much higher rewards while reducing constraint violations.
Authors: Till Freihaut, Giorgia Ramponi
Abstract: In multi-agent systems, the agent behavior is highly influenced by its utility function, as these utilities shape both individual goals as well as interactions with the other agents. Inverse Reinforcement Learning (IRL) is a well-established approach to inferring the utility function by observing an expert behavior within a given environment. In this paper, we extend the IRL framework to the multi-agent setting, assuming to observe agents who are following Nash Equilibrium (NE) policies. We theoretically investigate the set of utilities that explain the behavior of NE experts. Specifically, we provide an explicit characterization of the feasible reward set and analyze how errors in estimating the transition dynamics and expert behavior impact the recovered rewards. Building on these findings, we provide the first sample complexity analysis for the multi-agent IRL problem. Finally, we provide a numerical evaluation of our theoretical results.
Authors: Silin Zhou, Shuo Shang, Lisi Chen, Christian S. Jensen, Panos Kalnis
Abstract: Trajectory representation learning (TRL) maps trajectories to vectors that can then be used for various downstream tasks, including trajectory similarity computation, trajectory classification, and travel-time estimation. However, existing TRL methods often produce vectors that, when used in downstream tasks, yield insufficiently accurate results. A key reason is that they fail to utilize the comprehensive information encompassed by trajectories. We propose a self-supervised TRL framework, called RED, which effectively exploits multiple types of trajectory information. Overall, RED adopts the Transformer as the backbone model and masks the constituting paths in trajectories to train a masked autoencoder (MAE). In particular, RED considers the moving patterns of trajectories by employing a Road-aware masking strategy} that retains key paths of trajectories during masking, thereby preserving crucial information of the trajectories. RED also adopts a spatial-temporal-user joint Embedding scheme to encode comprehensive information when preparing the trajectories as model inputs. To conduct training, RED adopts Dual-objective task learning}: the Transformer encoder predicts the next segment in a trajectory, and the Transformer decoder reconstructs the entire trajectory. RED also considers the spatial-temporal correlations of trajectories by modifying the attention mechanism of the Transformer. We compare RED with 9 state-of-the-art TRL methods for 4 downstream tasks on 3 real-world datasets, finding that RED can usually improve the accuracy of the best-performing baseline by over 5%.
Authors: Arvind Mohan, Ashesh Chattopadhyay, Jonah Miller
Abstract: Differentiable Programming for scientific machine learning (SciML) has recently seen considerable interest and success, as it directly embeds neural networks inside PDEs, often called as NeuralPDEs, derived from first principle physics. Therefore, there is a widespread assumption in the community that NeuralPDEs are more trustworthy and generalizable than black box models. However, like any SciML model, differentiable programming relies predominantly on high-quality PDE simulations as "ground truth" for training. However, mathematics dictates that these are only discrete numerical approximations of the true physics. Therefore, we ask: Are NeuralPDEs and differentiable programming models trained on PDE simulations as physically interpretable as we think? In this work, we rigorously attempt to answer these questions, using established ideas from numerical analysis, experiments, and analysis of model Jacobians. Our study shows that NeuralPDEs learn the artifacts in the simulation training data arising from the discretized Taylor Series truncation error of the spatial derivatives. Additionally, NeuralPDE models are systematically biased, and their generalization capability is likely enabled by a fortuitous interplay of numerical dissipation and truncation error in the training dataset and NeuralPDE, which seldom happens in practical applications. This bias manifests aggressively even in relatively accessible 1-D equations, raising concerns about the veracity of differentiable programming on complex, high-dimensional, real-world PDEs, and in dataset integrity of foundation models. Further, we observe that the initial condition constrains the truncation error in initial-value problems in PDEs, thereby exerting limitations to extrapolation. Finally, we demonstrate that an eigenanalysis of model weights can indicate a priori if the model will be inaccurate for out-of-distribution testing.
Authors: Fengyuan Liu, Nikhil Kandpal, Colin Raffel
Abstract: The influence of contextual input on the behavior of large language models (LLMs) has prompted the development of context attribution methods that aim to quantify each context span's effect on an LLM's generations. The leave-one-out (LOO) error, which measures the change in the likelihood of the LLM's response when a given span of the context is removed, provides a principled way to perform context attribution, but can be prohibitively expensive to compute for large models. In this work, we introduce AttriBoT, a series of novel techniques for efficiently computing an approximation of the LOO error for context attribution. Specifically, AttriBoT uses cached activations to avoid redundant operations, performs hierarchical attribution to reduce computation, and emulates the behavior of large target models with smaller proxy models. Taken together, AttriBoT can provide a >300x speedup while remaining more faithful to a target model's LOO error than prior context attribution methods. This stark increase in performance makes computing context attributions for a given response 30x faster than generating the response itself, empowering real-world applications that require computing attributions at scale. We release a user-friendly and efficient implementation of AttriBoT to enable efficient LLM interpretability as well as encourage future development of efficient context attribution methods.
Authors: Valentino Delle Rose, Alexander Kozachinskiy, Tomasz Steifer
Abstract: Delle Rose et al.~(COLT'23) introduced an effective version of the Vapnik-Chervonenkis dimension, and showed that it characterizes improper PAC learning with total computable learners. In this paper, we introduce and study a similar effectivization of the notion of Littlestone dimension. Finite effective Littlestone dimension is a necessary condition for computable online learning but is not a sufficient one -- which we already establish for classes of the effective Littlestone dimension 2. However, the effective Littlestone dimension equals the optimal mistake bound for computable learners in two special cases: a) for classes of Littlestone dimension 1 and b) when the learner receives as additional information an upper bound on the numbers to be guessed. Interestingly, finite effective Littlestone dimension also guarantees that the class consists only of computable functions.
Authors: Hjalmar Wijk, Tao Lin, Joel Becker, Sami Jawhar, Neev Parikh, Thomas Broadley, Lawrence Chan, Michael Chen, Josh Clymer, Jai Dhyani, Elena Ericheva, Katharyn Garcia, Brian Goodrich, Nikola Jurkovic, Megan Kinniment, Aron Lajko, Seraphina Nix, Lucas Sato, William Saunders, Maksym Taran, Ben West, Elizabeth Barnes
Abstract: Frontier AI safety policies highlight automation of AI research and development (R&D) by AI agents as an important capability to anticipate. However, there exist few evaluations for AI R&D capabilities, and none that are highly realistic and have a direct comparison to human performance. We introduce RE-Bench (Research Engineering Benchmark, v1), which consists of 7 challenging, open-ended ML research engineering environments and data from 71 8-hour attempts by 61 distinct human experts. We confirm that our experts make progress in the environments given 8 hours, with 82% of expert attempts achieving a non-zero score and 24% matching or exceeding our strong reference solutions. We compare humans to several public frontier models through best-of-k with varying time budgets and agent designs, and find that the best AI agents achieve a score 4x higher than human experts when both are given a total time budget of 2 hours per environment. However, humans currently display better returns to increasing time budgets, narrowly exceeding the top AI agent scores given an 8-hour budget, and achieving 2x the score of the top AI agent when both are given 32 total hours (across different attempts). Qualitatively, we find that modern AI agents possess significant expertise in many ML topics -- e.g. an agent wrote a faster custom Triton kernel than any of our human experts' -- and can generate and test solutions over ten times faster than humans, at much lower cost. We open-source the evaluation environments, human expert data, analysis code and agent trajectories to facilitate future research.
Authors: Arnav M. Das, Chi Ian Tang, Fahim Kawsar, Mohammad Malekzadeh
Abstract: Sensing human motions through Inertial Measurement Units (IMUs) embedded in personal devices has enabled significant applications in health and wellness. While labeled IMU data is scarce, we can collect unlabeled or weakly labeled IMU data to model human motions. For video or text modalities, the "pretrain and adapt" approach utilizes large volumes of unlabeled or weakly labeled data for pretraining, building a strong feature extractor, followed by adaptation to specific tasks using limited labeled data. This approach has not been widely adopted in the IMU domain for two reasons: (1) pretraining methods are poorly understood in the context of IMU, and (2) open-source pretrained models that generalize across datasets are rarely publicly available. In this paper, we aim to address the first issue by proposing PRIMUS, a method for PRetraining IMU encoderS. We conduct a systematic and unified evaluation of various self-supervised and multimodal learning pretraining objectives. Our findings indicate that using PRIMUS, which combines self-supervision, multimodal supervision, and nearest-neighbor supervision, can significantly enhance downstream performance. With fewer than 500 labeled samples per class, PRIMUS effectively enhances downstream performance by up to 15% in held-out test data, compared to the state-of-the-art multimodal training method. To benefit the broader community, our code and pre-trained IMU encoders will be made publicly available at github.com/nokia-bell-labs upon publication.
Authors: Atilla P. Kiraly, Sebastien Baur, Kenneth Philbrick, Fereshteh Mahvar, Liron Yatziv, Tiffany Chen, Bram Sterling, Nick George, Fayaz Jamil, Jing Tang, Kai Bailey, Faruk Ahmed, Akshay Goel, Abbi Ward, Lin Yang, Andrew Sellergren, Yossi Matias, Avinatan Hassidim, Shravya Shetty, Daniel Golden, Shekoofeh Azizi, David F. Steiner, Yun Liu, Tim Thelin, Rory Pilgrim, Can Kirmizibayrak
Abstract: Robust medical Machine Learning (ML) models have the potential to revolutionize healthcare by accelerating clinical research, improving workflows and outcomes, and producing novel insights or capabilities. Developing such ML models from scratch is cost prohibitive and requires substantial compute, data, and time (e.g., expert labeling). To address these challenges, we introduce Health AI Developer Foundations (HAI-DEF), a suite of pre-trained, domain-specific foundation models, tools, and recipes to accelerate building ML for health applications. The models cover various modalities and domains, including radiology (X-rays and computed tomography), histopathology, dermatological imaging, and audio. These models provide domain specific embeddings that facilitate AI development with less labeled data, shorter training times, and reduced computational costs compared to traditional approaches. In addition, we utilize a common interface and style across these models, and prioritize usability to enable developers to integrate HAI-DEF efficiently. We present model evaluations across various tasks and conclude with a discussion of their application and evaluation, covering the importance of ensuring efficacy, fairness, and equity. Finally, while HAI-DEF and specifically the foundation models lower the barrier to entry for ML in healthcare, we emphasize the importance of validation with problem- and population-specific data for each desired usage setting. This technical report will be updated over time as more modalities and features are added.
Authors: David J Poland, Lemuel Puglisi, Daniele Ravi
Abstract: This article introduces Transformer Quantile Regression Neural Networks (TQRNNs), a novel data-driven solution for real-time machine failure prediction in manufacturing contexts. Our objective is to develop an advanced predictive maintenance model capable of accurately identifying machine system breakdowns. To do so, TQRNNs employ a two-step approach: (i) a modified quantile regression neural network to segment anomaly outliers while maintaining low time complexity, and (ii) a concatenated transformer network aimed at facilitating accurate classification even within a large timeframe of up to one hour. We have implemented our proposed pipeline in a real-world beverage manufacturing industry setting. Our findings demonstrate the model's effectiveness, achieving an accuracy rate of 70.84% with a 1-hour lead time for predicting machine breakdowns. Additionally, our analysis shows that using TQRNNs can increase high-quality production, improving product yield from 78.38% to 89.62%. We believe that predictive maintenance assumes a pivotal role in modern manufacturing, minimizing unplanned downtime, reducing repair costs, optimizing production efficiency, and ensuring operational stability. Its potential to generate substantial cost savings while enhancing sustainability and competitiveness underscores its importance in contemporary manufacturing practices.
Authors: Marco Fiandri, Alberto Maria Metelli, Francesco Trov`o
Abstract: This paper is in the field of stochastic Multi-Armed Bandits (MABs), i.e. those sequential selection techniques able to learn online using only the feedback given by the chosen option (a.k.a. $arm$). We study a particular case of the rested bandits in which the arms' expected reward is monotonically non-decreasing and concave. We study the inherent sample complexity of the regret minimization problem by deriving suitable regret lower bounds. Then, we design an algorithm for the rested case $\textit{R-ed-UCB}$, providing a regret bound depending on the properties of the instance and, under certain circumstances, of $\widetilde{\mathcal{O}}(T^{\frac{2}{3}})$. We empirically compare our algorithms with state-of-the-art methods for non-stationary MABs over several synthetically generated tasks and an online model selection problem for a real-world dataset
Authors: Jeongjin Shin, Sangdon Park
Abstract: Deep learning models are vulnerable to backdoor attacks, where adversaries inject malicious functionality during training that activates on trigger inputs at inference time. Extensive research has focused on developing stealthy backdoor attacks to evade detection and defense mechanisms. However, these approaches still have limitations that leave the door open for detection and mitigation due to their inherent design to cause malicious behavior in the presence of a trigger. To address this limitation, we introduce Deferred Backdoor Functionality Activation (DBFA), a new paradigm in backdoor attacks. Unlike conventional attacks, DBFA initially conceals its backdoor, producing benign outputs even when triggered. This stealthy behavior allows DBFA to bypass multiple detection and defense methods, remaining undetected during initial inspections. The backdoor functionality is strategically activated only after the model undergoes subsequent updates, such as retraining on benign data. DBFA attacks exploit the common practice in the life cycle of machine learning models to perform model updates and fine-tuning after initial deployment. To implement DBFA attacks, we approach the problem by making the unlearning of the backdoor fragile, allowing it to be easily cancelled and subsequently reactivate the backdoor functionality. To achieve this, we propose a novel two-stage training scheme, called DeferBad. Our extensive experiments across various fine-tuning scenarios, backdoor attack types, datasets, and model architectures demonstrate the effectiveness and stealthiness of DeferBad.
Authors: Harish Haresamudram, Chi Ian Tang, Sungho Suh, Paul Lukowicz, Thomas Ploetz
Abstract: In the many years since the inception of wearable sensor-based Human Activity Recognition (HAR), a wide variety of methods have been introduced and evaluated for their ability to recognize activities. Substantial gains have been made since the days of hand-crafting heuristics as features, yet, progress has seemingly stalled on many popular benchmarks, with performance falling short of what may be considered 'sufficient'-- despite the increase in computational power and scale of sensor data, as well as rising complexity in techniques being employed. The HAR community approaches a new paradigm shift, this time incorporating world knowledge from foundational models. In this paper, we take stock of sensor-based HAR -- surveying it from its beginnings to the current state of the field, and charting its future. This is accompanied by a hands-on tutorial, through which we guide practitioners in developing HAR systems for real-world application scenarios. We provide a compendium for novices and experts alike, of methods that aim at finally solving the activity recognition problem.
Authors: Palak (Microsoft Research India), Rohan Gandhi (Microsoft Research India), Karan Tandon (Microsoft Research India), Debopam Bhattacherjee (Microsoft Research India), Venkata N. Padmanabhan (Microsoft Research India)
Abstract: The widespread adoption of language models (LMs) across multiple industries has caused huge surge in demand for GPUs. Training LMs requires tens of thousands of GPUs and housing them in the same datacenter (DCs) is becoming challenging. We focus on training such models across multiple DCs connected via Wide-Area-Network (WAN). We build ATLAS that speeds up such training time using novel temporal bandwidth sharing and many other design choices. While ATLAS improves the training time, it does not eliminate the bubbles (idle GPU cycles). We built BUBBLETEA that runs prefill-as-a-service (part of LM inference) during the bubbles that improves the GPU utilization substantially without any impact of training. Together, ATLAS and BUBBLETEA improve training time by up to 17X and achieve GPU utilization of up to 94%.
Authors: Yao Xu, Shizhu He, Zeng Xiangrong, Jiabei Chen, Guang Liu, Bingning Wang, Jun Zhao, Kang Liu
Abstract: Structured data, such as tables, graphs, and databases, play a critical role in plentiful NLP tasks such as question answering and dialogue system. Recently, inspired by Vision-Language Models, Graph Neutral Networks (GNNs) have been introduced as an additional modality into the input of Large Language Models (LLMs) to improve their performance on Structured Knowledge Grounding (SKG) tasks. However, those GNN-enhanced LLMs have the following limitations: (1) They employ diverse GNNs to model varying types of structured data, rendering them unable to uniformly process various forms of structured data. (2) The pretraining of GNNs is coupled with specific LLMs, which prevents GNNs from fully aligning with the textual space and limits their adaptability to other LLMs. To address these issues, we propose \textbf{L}arge \textbf{L}anguage and \textbf{S}tructured Data \textbf{A}ssistant (LLaSA), a general framework for enhancing LLMs' ability to handle structured data. Specifically, we represent various types of structured data in a unified hypergraph format, and use self-supervised learning to pretrain a hypergraph encoder, and a G-Former compressing encoded hypergraph representations with cross-attention. The compressed hypergraph representations are appended to the serialized inputs during training and inference stages of LLMs. Experimental results on multiple SKG tasks show that our pretrained hypergraph encoder can adapt to various LLMs and enhance their ability to process different types of structured data. Besides, LLaSA, with LoRA fine-tuning, outperforms previous SOTA method using full parameters tuning.
Authors: Apurva Kalia, Dilip Krishnan, Soha Hassoun
Abstract: Motivation: A major challenge in metabolomics is annotation: assigning molecular structures to mass spectral fragmentation patterns. Despite recent advances in molecule-to-spectra and in spectra-to-molecular fingerprint prediction (FP), annotation rates remain low. Results: We introduce in this paper a novel paradigm (JESTR) for annotation. Unlike prior approaches that explicitly construct molecular fingerprints or spectra, JESTR leverages the insight that molecules and their corresponding spectra are views of the same data and effectively embeds their representations in a joint space. Candidate structures are ranked based on cosine similarity between the embeddings of query spectrum and each candidate. We evaluate JESTR against mol-to-spec and spec-to-FP annotation tools on three datasets. On average, for rank@[1-5], JESTR outperforms other tools by 23.6%-71.6%. We further demonstrate the strong value of regularization with candidate molecules during training, boosting rank@1 performance by 11.4% and enhancing the model's ability to discern between target and candidate molecules. Through JESTR, we offer a novel promising avenue towards accurate annotation, therefore unlocking valuable insights into the metabolome.
Authors: Elizaveta Reganova, Peter Steinbach
Abstract: Large Language Models (LLMs) have gained significant popularity in recent years for their ability to answer questions in various fields. However, these models have a tendency to "hallucinate" their responses, making it challenging to evaluate their performance. A major challenge is determining how to assess the certainty of a model's predictions and how it correlates with accuracy. In this work, we introduce an analysis for evaluating the performance of popular open-source LLMs, as well as gpt-3.5 Turbo, on multiple choice physics questionnaires. We focus on the relationship between answer accuracy and variability in topics related to physics. Our findings suggest that most models provide accurate replies in cases where they are certain, but this is by far not a general behavior. The relationship between accuracy and uncertainty exposes a broad horizontal bell-shaped distribution. We report how the asymmetry between accuracy and uncertainty intensifies as the questions demand more logical reasoning of the LLM agent, while the same relationship remains sharp for knowledge retrieval tasks.
Authors: Ross Gardiner, Sareh Rowands, Benno I. Simmons
Abstract: Camera traps, combined with AI, have emerged as a way to achieve automated, scalable biodiversity monitoring. However, the passive infrared (PIR) sensors that trigger camera traps are poorly suited for detecting small, fast-moving ectotherms such as insects. Insects comprise over half of all animal species and are key components of ecosystems and agriculture. The need for an appropriate and scalable insect camera trap is critical in the wake of concerning reports of declines in insect populations. This study proposes an alternative to the PIR trigger: ultra-lightweight convolutional neural networks running on low-powered hardware to detect insects in a continuous stream of captured images. We train a suite of models to distinguish insect images from backgrounds. Our design achieves zero latency between trigger and image capture. Our models are rigorously tested and achieve high accuracy ranging from 91.8% to 96.4% AUC on validation data and >87% AUC on data from distributions unseen during training. The high specificity of our models ensures minimal saving of false positive images, maximising deployment storage efficiency. High recall scores indicate a minimal false negative rate, maximising insect detection. Further analysis with saliency maps shows the learned representation of our models to be robust, with low reliance on spurious background features. Our system is also shown to operate deployed on off-the-shelf, low-powered microcontroller units, consuming a maximum power draw of less than 300mW. This enables longer deployment times using cheap and readily available battery components. Overall we offer a step change in the cost, efficiency and scope of insect monitoring. Solving the challenging trigger problem, we demonstrate a system which can be deployed for far longer than existing designs and budgets power and bandwidth effectively, moving towards a generic insect camera trap.
Authors: Aditya Sridhar
Abstract: Music genre classification is a critical component of music recommendation systems, generation algorithms, and cultural analytics. In this work, we present an innovative model for classifying music genres using attention-based temporal signature modeling. By processing spectrogram sequences through Convolutional Neural Networks (CNNs) and multi-head attention layers, our approach captures the most temporally significant moments within each piece, crafting a unique "signature" for genre identification. This temporal focus not only enhances classification accuracy but also reveals insights into genre-specific characteristics that can be intuitively mapped to listener perceptions. Our findings offer potential applications in personalized music recommendation systems by highlighting cross-genre similarities and distinctiveness, aligning closely with human musical intuition. This work bridges the gap between technical classification tasks and the nuanced, human experience of genre.
Authors: Jingtao Ding, Yunke Zhang, Yu Shang, Yuheng Zhang, Zefang Zong, Jie Feng, Yuan Yuan, Hongyuan Su, Nian Li, Nicholas Sukiennik, Fengli Xu, Yong Li
Abstract: The concept of world models has garnered significant attention due to advancements in multimodal large language models such as GPT-4 and video generation models such as Sora, which are central to the pursuit of artificial general intelligence. This survey offers a comprehensive review of the literature on world models. Generally, world models are regarded as tools for either understanding the present state of the world or predicting its future dynamics. This review presents a systematic categorization of world models, emphasizing two primary functions: (1) constructing internal representations to understand the mechanisms of the world, and (2) predicting future states to simulate and guide decision-making. Initially, we examine the current progress in these two categories. We then explore the application of world models in key domains, including autonomous driving, robotics, and social simulacra, with a focus on how each domain utilizes these aspects. Finally, we outline key challenges and provide insights into potential future research directions.
Authors: Qingquan Zhang, Qiqi Duan, Bo Yuan, Yuhui Shi, Jialin Liu
Abstract: Large Language Models (LLMs) have made significant strides in the field of artificial intelligence, showcasing their ability to interact with humans and influence human cognition through information dissemination. However, recent studies have brought to light instances of bias inherent within these LLMs, presenting a critical issue that demands attention. In our research, we delve deeper into the intricate challenge of harmonising accuracy and fairness in the enhancement of LLMs. While improving accuracy can indeed enhance overall LLM performance, it often occurs at the expense of fairness. Overemphasising optimisation of one metric invariably leads to a significant degradation of the other. This underscores the necessity of taking into account multiple considerations during the design and optimisation phases of LLMs. Therefore, we advocate for reformulating the LLM training process as a multi-objective learning task. Our investigation reveals that multi-objective evolutionary learning (MOEL) methodologies offer promising avenues for tackling this challenge. Our MOEL framework enables the simultaneous optimisation of both accuracy and fairness metrics, resulting in a Pareto-optimal set of LLMs. In summary, our study sheds valuable lights on the delicate equilibrium between accuracy and fairness within LLMs, which is increasingly significant for their real-world applications. By harnessing MOEL, we present a promising pathway towards fairer and more efficacious AI technologies.
Authors: Mohammad Arafat Ullah, Arthy Anjum, Rashedul Amin Tuhin, Shamim Akhter
Abstract: A distributed denial-of-service (DDoS) attack is an attempt to produce humongous traffic within a network by overwhelming a targeted server or its neighboring infrastructure with a flood of service requests ceaselessly coming from multiple remotely controlled malware-infected computers or network-connected devices. Thus, exploring DDoS attacks by recognizing their functionalities and differentiating them from normal traffic services are the primary concerns of network security issues particularly for online businesses. In modern networks, most DDoS attacks occur in the network and application layer including HTTP flood, UDP flood, SIDDOS, SMURF, SNMP flood, IP NULL, etc. The goal of this paper is to detect DDoS attacks from all service requests and classify them according to DDoS classes. In this regard, a standard dataset is collected from the internet which contains several network-related attributes and their corresponding DDoS attack class name. Two(2) different machine learning approaches, SVM and Logistic Regression, are implemented in the dataset for detecting and classifying DDoS attacks, and a comparative study is accomplished among them in terms of accuracy, precision, and recall rates. Logistic Regression and SVM both achieve 98.65% classification accuracy which is the highest achieved accuracy among other previous experiments with the same dataset.
Authors: Eden Luzon, Guy Amit, Roy Weiss, Yisroel Mirsky
Abstract: Neural networks, such as image classifiers, are frequently trained on proprietary and confidential datasets. It is generally assumed that once deployed, the training data remains secure, as adversaries are limited to query response interactions with the model, where at best, fragments of arbitrary data can be inferred without any guarantees on their authenticity. In this paper, we propose the memory backdoor attack, where a model is covertly trained to memorize specific training samples and later selectively output them when triggered with an index pattern. What makes this attack unique is that it (1) works even when the tasks conflict (making a classifier output images), (2) enables the systematic extraction of training samples from deployed models and (3) offers guarantees on the extracted authenticity of the data. We demonstrate the attack on image classifiers, segmentation models, and a large language model (LLM). We demonstrate the attack on image classifiers, segmentation models, and a large language model (LLM). With this attack, it is possible to hide thousands of images and texts in modern vision architectures and LLMs respectively, all while maintaining model performance. The memory back door attack poses a significant threat not only to conventional model deployments but also to federated learning paradigms and other modern frameworks. Therefore, we suggest an efficient and effective countermeasure that can be immediately applied and advocate for further work on the topic.
Authors: Arthur Elwing Torres, Edleno Silva de Moura, Altigran Soares da Silva, Mario A. Nascimento, Filipe Mesquita
Abstract: Named Entity Recognition (NER) is a machine learning task that traditionally relies on supervised learning and annotated data. Acquiring such data is often a challenge, particularly in specialized fields like medical, legal, and financial sectors. Those are commonly referred to as low-resource domains, which comprise long-tail entities, due to the scarcity of available data. To address this, data augmentation techniques are increasingly being employed to generate additional training instances from the original dataset. In this study, we evaluate the effectiveness of two prominent text augmentation techniques, Mention Replacement and Contextual Word Replacement, on two widely-used NER models, Bi-LSTM+CRF and BERT. We conduct experiments on four datasets from low-resource domains, and we explore the impact of various combinations of training subset sizes and number of augmented examples. We not only confirm that data augmentation is particularly beneficial for smaller datasets, but we also demonstrate that there is no universally optimal number of augmented examples, i.e., NER practitioners must experiment with different quantities in order to fine-tune their projects.
Authors: Selma Husanovic, Ginger Egberts, Alexander Heinlein, Fred Vermolen
Abstract: Burn injuries present a significant global health challenge. Among the most severe long-term consequences are contractures, which can lead to functional impairments and disfigurement. Understanding and predicting the evolution of post-burn wounds is essential for developing effective treatment strategies. Traditional mathematical models, while accurate, are often computationally expensive and time-consuming, limiting their practical application. Recent advancements in machine learning, particularly in deep learning, offer promising alternatives for accelerating these predictions. This study explores the use of a deep operator network (DeepONet), a type of neural operator, as a surrogate model for finite element simulations, aimed at predicting post-burn contraction across multiple wound shapes. A DeepONet was trained on three distinct initial wound shapes, with enhancement made to the architecture by incorporating initial wound shape information and applying sine augmentation to enforce boundary conditions. The performance of the trained DeepONet was evaluated on a test set including finite element simulations based on convex combinations of the three basic wound shapes. The model achieved an $R^2$ score of $0.99$, indicating strong predictive accuracy and generalization. Moreover, the model provided reliable predictions over an extended period of up to one year, with speedups of up to 128-fold on CPU and 235-fold on GPU, compared to the numerical model. These findings suggest that DeepONets can effectively serve as a surrogate for traditional finite element methods in simulating post-burn wound evolution, with potential applications in medical treatment planning.
Authors: Sizhe Wang, Wenwen Li
Abstract: This study introduces a novel approach to terrain feature classification by incorporating spatial point pattern statistics into deep learning models. Inspired by the concept of location encoding, which aims to capture location characteristics to enhance GeoAI decision-making capabilities, we improve the GeoAI model by a knowledge driven approach to integrate both first-order and second-order effects of point patterns. This paper investigates how these spatial contexts impact the accuracy of terrain feature predictions. The results show that incorporating spatial point pattern statistics notably enhances model performance by leveraging different representations of spatial relationships.
Authors: Jing Liu, Yang Liu, Xiaoguang Zhu
Abstract: Video Anomaly Detection (VAD) aims to automatically analyze spatiotemporal patterns in surveillance videos collected from open spaces to detect anomalous events that may cause harm without physical contact. However, vision-based surveillance systems such as closed-circuit television often capture personally identifiable information. The lack of transparency and interpretability in video transmission and usage raises public concerns about privacy and ethics, limiting the real-world application of VAD. Recently, researchers have focused on privacy concerns in VAD by conducting systematic studies from various perspectives including data, features, and systems, making Privacy-Preserving Video Anomaly Detection (P2VAD) a hotspot in the AI community. However, current research in P2VAD is fragmented, and prior reviews have mostly focused on methods using RGB sequences, overlooking privacy leakage and appearance bias considerations. To address this gap, this article systematically reviews the progress of P2VAD for the first time, defining its scope and providing an intuitive taxonomy. We outline the basic assumptions, learning frameworks, and optimization objectives of various approaches, analyzing their strengths, weaknesses, and potential correlations. Additionally, we provide open access to research resources such as benchmark datasets and available code. Finally, we discuss key challenges and future opportunities from the perspectives of AI development and P2VAD deployment, aiming to guide future work in the field.
Authors: Chunwei Liu, Enrique Noriega-Atala, Adarsh Pyarelal, Clayton T Morrison, Mike Cafarella
Abstract: The global output of academic publications exceeds 5 million articles per year, making it difficult for humans to keep up with even a tiny fraction of scientific output. We need methods to navigate and interpret the artifacts -- texts, graphs, charts, code, models, and datasets -- that make up the literature. This paper evaluates various methods for extracting mathematical model variables from epidemiological studies, such as ``infection rate ($\alpha$),'' ``recovery rate ($\gamma$),'' and ``mortality rate ($\mu$).'' Variable extraction appears to be a basic task, but plays a pivotal role in recovering models from scientific literature. Once extracted, we can use these variables for automatic mathematical modeling, simulation, and replication of published results. We introduce a benchmark dataset comprising manually-annotated variable descriptions and variable values extracted from scientific papers. Based on this dataset, we present several baseline methods for variable extraction based on Large Language Models (LLMs) and rule-based information extraction systems. Our analysis shows that LLM-based solutions perform the best. Despite the incremental benefits of combining rule-based extraction outputs with LLMs, the leap in performance attributed to the transfer-learning and instruction-tuning capabilities of LLMs themselves is far more significant. This investigation demonstrates the potential of LLMs to enhance automatic comprehension of scientific artifacts and for automatic model recovery and simulation.
Authors: Larry Schester, Luis E. Ortiz
Abstract: Vehicles today can drive themselves on highways and driverless robotaxis operate in major cities, with more sophisticated levels of autonomous driving expected to be available and become more common in the future. Yet, technically speaking, so-called "Level 5" (L5) operation, corresponding to full autonomy, has not been achieved. For that to happen, functions such as fully autonomous highway ramp entry must be available, and provide provably safe, and reliably robust behavior to enable full autonomy. We present a systematic study of a highway ramp function that controls the vehicles forward-moving actions to minimize collisions with the stream of highway traffic into which a merging (ego) vehicle enters. We take a game-theoretic multi-agent (MA) approach to this problem and study the use of controllers based on deep reinforcement learning (DRL). The virtual environment of the MA DRL uses self-play with simulated data where merging vehicles safely learn to control longitudinal position during a taper-type merge. The work presented in this paper extends existing work by studying the interaction of more than two vehicles (agents) and does so by systematically expanding the road scene with additional traffic and ego vehicles. While previous work on the two-vehicle setting established that collision-free controllers are theoretically impossible in fully decentralized, non-coordinated environments, we empirically show that controllers learned using our approach are nearly ideal when measured against idealized optimal controllers.
Authors: Dmitry Kovalev, Ekaterina Borodich
Abstract: We revisit the smooth convex-concave bilinearly-coupled saddle-point problem of the form $\min_x\max_y f(x) + \langle y,\mathbf{B} x\rangle - g(y)$. In the highly specific case where each of the functions $f(x)$ and $g(y)$ is either affine or strongly convex, there exist lower bounds on the number of gradient evaluations and matrix-vector multiplications required to solve the problem, as well as matching optimal algorithms. A notable aspect of these algorithms is that they are able to attain linear convergence, i.e., the number of iterations required to solve the problem is proportional to $\log(1/\epsilon)$. However, the class of bilinearly-coupled saddle-point problems for which linear convergence is possible is much wider and can involve smooth non-strongly convex functions $f(x)$ and $g(y)$. Therefore, we develop the first lower complexity bounds and matching optimal linearly converging algorithms for this problem class. Our lower complexity bounds are much more general, but they cover and unify the existing results in the literature. On the other hand, our algorithm implements the separation of complexities, which, for the first time, enables the simultaneous achievement of both optimal gradient evaluation and matrix-vector multiplication complexities, resulting in the best theoretical performance to date.
Authors: Alex Mathai, Kranthi Sedamaki, Debeshee Das, Noble Saji Mathews, Srikanth Tamilselvam, Sridhar Chimalakonda, Atul Kumar
Abstract: Machine Learning (ML) for software engineering (SE) has gained prominence due to its ability to significantly enhance the performance of various SE applications. This progress is largely attributed to the development of generalizable source code representations that effectively capture the syntactic and semantic characteristics of code. In recent years, pre-trained transformer-based models, inspired by natural language processing (NLP), have shown remarkable success in SE tasks. However, source code contains structural and semantic properties embedded within its grammar, which can be extracted from structured code-views like the Abstract Syntax Tree (AST), Data-Flow Graph (DFG), and Control-Flow Graph (CFG). These code-views can complement NLP techniques, further improving SE tasks. Unfortunately, there are no flexible frameworks to infuse arbitrary code-views into existing transformer-based models effectively. Therefore, in this work, we propose CodeSAM, a novel scalable framework to infuse multiple code-views into transformer-based models by creating self-attention masks. We use CodeSAM to fine-tune a small language model (SLM) like CodeBERT on the downstream SE tasks of semantic code search, code clone detection, and program classification. Experimental results show that by using this technique, we improve downstream performance when compared to SLMs like GraphCodeBERT and CodeBERT on all three tasks by utilizing individual code-views or a combination of code-views during fine-tuning. We believe that these results are indicative that techniques like CodeSAM can help create compact yet performant code SLMs that fit in resource constrained settings.
Authors: Hao-Wen Dong
Abstract: Generative AI has been transforming the way we interact with technology and consume content. In the next decade, AI technology will reshape how we create audio content in various media, including music, theater, films, games, podcasts, and short videos. In this dissertation, I introduce the three main directions of my research centered around generative AI for music and audio: 1) multitrack music generation, 2) assistive music creation tools, and 3) multimodal learning for audio and music. Through my research, I aim to answer the following two fundamental questions: 1) How can AI help professionals or amateurs create music and audio content? 2) Can AI learn to create music in a way similar to how humans learn music? My long-term goal is to lower the barrier of entry for music composition and democratize audio content creation
Authors: Zimo Wang, Cheng Wang, Taiki Yoshino, Sirui Tao, Ziyang Fu, Tzu-Mao Li
Abstract: We propose a method, HotSpot, for optimizing neural signed distance functions, based on a relation between the solution of a screened Poisson equation and the distance function. Existing losses such as the eikonal loss cannot guarantee the recovered implicit function to be a distance function, even when the implicit function satisfies the eikonal equation almost everywhere. Furthermore, the eikonal loss suffers from stability issues in optimization and the remedies that introduce area or divergence minimization can lead to oversmoothing. We address these challenges by designing a loss function that when minimized can converge to the true distance function, is stable, and naturally penalize large surface area. We provide theoretical analysis and experiments on both challenging 2D and 3D datasets and show that our method provide better surface reconstruction and more accurate distance approximation.
Authors: Mengze Gao, Zachary Shah, Xiaozhi Cao, Nan Wang, Daniel Abraham, Kawin Setsompop
Abstract: Spatiotemporal magnetic field variations from B0-inhomogeneity and diffusion-encoding-induced eddy-currents can be detrimental to rapid image-encoding schemes such as spiral, EPI and 3D-cones, resulting in undesirable image artifacts. In this work, a data driven approach for automatic estimation of these field imperfections is developed by combining autofocus metrics with deep learning, and by leveraging a compact basis representation of the expected field imperfections. The method was applied to single-shot spiral diffusion MRI at high b-values where accurate estimation of B0 and eddy were obtained, resulting in high quality image reconstruction without need for additional external calibrations.
Authors: Pura Peetathawatchai, Wei-Ning Chen, Berivan Isik, Sanmi Koyejo, Albert No
Abstract: We introduce novel methods for adapting diffusion models under differential privacy (DP) constraints, enabling privacy-preserving style and content transfer without fine-tuning. Traditional approaches to private adaptation, such as DP-SGD, incur significant computational overhead and degrade model performance when applied to large, complex models. Our approach instead leverages embedding-based techniques: Universal Guidance and Textual Inversion (TI), adapted with differentially private mechanisms. We apply these methods to Stable Diffusion for style adaptation using two private datasets: a collection of artworks by a single artist and pictograms from the Paris 2024 Olympics. Experimental results show that the TI-based adaptation achieves superior fidelity in style transfer, even under strong privacy guarantees, while both methods maintain high privacy resilience by employing calibrated noise and subsampling strategies. Our findings demonstrate a feasible and efficient pathway for privacy-preserving diffusion model adaptation, balancing data protection with the fidelity of generated images, and offer insights into embedding-driven methods for DP in generative AI applications.
Authors: Anindya De, Shivam Nadimpalli, Ryan O'Donnell, Rocco A. Servedio
Abstract: We give a dimension-independent sparsification result for suprema of centered Gaussian processes: Let $T$ be any (possibly infinite) bounded set of vectors in $\mathbb{R}^n$, and let $\{{\boldsymbol{X}}_t\}_{t\in T}$ be the canonical Gaussian process on $T$. We show that there is an $O_\varepsilon(1)$-size subset $S \subseteq T$ and a set of real values $\{c_s\}_{s \in S}$ such that $\sup_{s \in S} \{{\boldsymbol{X}}_s + c_s\}$ is an $\varepsilon$-approximator of $\sup_{t \in T} {\boldsymbol{X}}_t$. Notably, the size of $S$ is completely independent of both the size of $T$ and of the ambient dimension $n$. We use this to show that every norm is essentially a junta when viewed as a function over Gaussian space: Given any norm $\nu(x)$ on $\mathbb{R}^n$, there is another norm $\psi(x)$ which depends only on the projection of $x$ along $O_\varepsilon(1)$ directions, for which $\psi({\boldsymbol{g}})$ is a multiplicative $(1 \pm \varepsilon)$-approximation of $\nu({\boldsymbol{g}})$ with probability $1-\varepsilon$ for ${\boldsymbol{g}} \sim N(0,I_n)$. We also use our sparsification result for suprema of centered Gaussian processes to give a sparsification lemma for convex sets of bounded geometric width: Any intersection of (possibly infinitely many) halfspaces in $\mathbb{R}^n$ that are at distance $O(1)$ from the origin is $\varepsilon$-close, under $N(0,I_n)$, to an intersection of only $O_\varepsilon(1)$ many halfspaces. We describe applications to agnostic learning and tolerant property testing.
Authors: Lynda Aouar, Han Yu
Abstract: Adaptive causal representation learning from observational data is presented, integrated with an efficient sample splitting technique within the semiparametric estimating equation framework. The support points sample splitting (SPSS), a subsampling method based on energy distance, is employed for efficient double machine learning (DML) in causal inference. The support points are selected and split as optimal representative points of the full raw data in a random sample, in contrast to the traditional random splitting, and providing an optimal sub-representation of the underlying data generating distribution. They offer the best representation of a full big dataset, whereas the unit structural information of the underlying distribution via the traditional random data splitting is most likely not preserved. Three machine learning estimators were adopted for causal inference, support vector machine (SVM), deep learning (DL), and a hybrid super learner (SL) with deep learning (SDL), using SPSS. A comparative study is conducted between the proposed SVM, DL, and SDL representations using SPSS, and the benchmark results from Chernozhukov et al. (2018), which employed random forest, neural network, and regression trees with a random k-fold cross-fitting technique on the 401(k)-pension plan real data. The simulations show that DL with SPSS and the hybrid methods of DL and SL with SPSS outperform SVM with SPSS in terms of computational efficiency and the estimation quality, respectively.
Authors: Vedant Mehta
Abstract: Neurological and Physiological Disorders that impact emotional regulation each have their own unique characteristics which are important to understand in order to create a generalized solution to all of them. The purpose of this experiment is to explore the potential applications of EEG-based Brain-Computer Interfaces (BCIs) in enhancing emotional regulation for individuals with neurological and physiological disorders. The research focuses on the development of a novel neural network algorithm for understanding EEG data, with a particular emphasis on recognizing and regulating emotional states. The procedure involves the collection of EEG-based emotion data from open-Neuro. Using novel data modification techniques, information from the dataset can be altered to create a dataset that has neural patterns of patients with disorders whilst showing emotional change. The data analysis reveals promising results, as the algorithm is able to successfully classify emotional states with a high degree of accuracy. This suggests that EEG-based BCIs have the potential to be a valuable tool in aiding individuals with a range of neurological and physiological disorders in recognizing and regulating their emotions. To improve upon this work, data collection on patients with neurological disorders should be done to improve overall sample diversity.
Authors: Xiaoxu Tian, Daniel Holdaway, Daryl Kleist
Abstract: The use of machine learning (ML) models in meteorology has attracted significant attention for their potential to improve weather forecasting efficiency and accuracy. GraphCast and NeuralGCM, two promising ML-based weather models, are at the forefront of this innovation. However, their suitability for data assimilation (DA) systems, particularly for four-dimensional variational (4DVar) DA, remains under-explored. This study evaluates the tangent linear (TL) and adjoint (AD) models of both GraphCast and NeuralGCM to assess their viability for integration into a DA framework. We compare the TL/AD results of GraphCast and NeuralGCM with those of the Model for Prediction Across Scales - Atmosphere (MPAS-A), a well-established numerical weather prediction (NWP) model. The comparison focuses on the physical consistency and reliability of TL/AD responses to perturbations. While the adjoint results of both GraphCast and NeuralGCM show some similarity to those of MPAS-A, they also exhibit unphysical noise at various vertical levels, raising concerns about their robustness for operational DA systems. The implications of this study extend beyond 4DVar applications. Unphysical behavior and noise in ML-derived TL/AD models could lead to inaccurate error covariances and unreliable ensemble forecasts, potentially degrading the overall performance of ensemble-based DA systems, as well. Addressing these challenges is critical to ensuring that ML models, such as GraphCast and NeuralGCM, can be effectively integrated into operational DA systems, paving the way for more accurate and efficient weather predictions.
Authors: AJ Piergiovanni, Dahun Kim, Michael S. Ryoo, Isaac Noble, Anelia Angelova
Abstract: Generating automatic dense captions for videos that accurately describe their contents remains a challenging area of research. Most current models require processing the entire video at once. Instead, we propose an efficient, online approach which outputs frequent, detailed and temporally aligned captions, without access to future frames. Our model uses a novel autoregressive factorized decoding architecture, which models the sequence of visual features for each time segment, outputting localized descriptions and efficiently leverages the context from the previous video segments. This allows the model to output frequent, detailed captions to more comprehensively describe the video, according to its actual local content, rather than mimic the training data. Second, we propose an optimization for efficient training and inference, which enables scaling to longer videos. Our approach shows excellent performance compared to both offline and online methods, and uses 20\% less compute. The annotations produced are much more comprehensive and frequent, and can further be utilized in automatic video tagging and in large-scale video data harvesting.
Authors: Kedi Zheng, Hongye Guo, Qixin Chen
Abstract: This paper studies the pool strategy for price-makers under imperfect information. In this occasion, market participants cannot obtain essential transmission parameters of the power system. Thus, price-makers should estimate the market results with respect to their offer curves using available historical information. The linear programming model of economic dispatch is analyzed with the theory of rim multi-parametric linear programming (rim-MPLP). The characteristics of system patterns (combinations of status flags for generating units and transmission lines) are revealed. A multi-class classification model based on support vector machine (SVM) is trained to map the offer curves to system patterns, which is then integrated into the decision framework of the price-maker. The performance of the proposed method is validated on the IEEE 30-bus system, Illinois synthetic 200-bus system, and South Carolina synthetic 500-bus system.
Authors: Hao Chen, Francois Bremond, Nicu Sebe, Shiliang Zhang
Abstract: Regular unsupervised domain adaptive person re-identification (ReID) focuses on adapting a model from a source domain to a fixed target domain. However, an adapted ReID model can hardly retain previously-acquired knowledge and generalize to unseen data. In this paper, we propose a Dual-level Joint Adaptation and Anti-forgetting (DJAA) framework, which incrementally adapts a model to new domains without forgetting source domain and each adapted target domain. We explore the possibility of using prototype and instance-level consistency to mitigate the forgetting during the adaptation. Specifically, we store a small number of representative image samples and corresponding cluster prototypes in a memory buffer, which is updated at each adaptation step. With the buffered images and prototypes, we regularize the image-to-image similarity and image-to-prototype similarity to rehearse old knowledge. After the multi-step adaptation, the model is tested on all seen domains and several unseen domains to validate the generalization ability of our method. Extensive experiments demonstrate that our proposed method significantly improves the anti-forgetting, generalization and backward-compatible ability of an unsupervised person ReID model.
Authors: Jinglei Cheng, Ruilin Zhou, Yuhang Gan, Chen Qian, Junyu Liu
Abstract: We introduce Quantum Hamiltonian Descent as a novel approach to solve the graph partition problem. By reformulating graph partition as a Quadratic Unconstrained Binary Optimization (QUBO) problem, we leverage QHD's quantum-inspired dynamics to identify optimal community structures. Our method implements a multi-level refinement strategy that alternates between QUBO formulation and QHD optimization to iteratively improve partition quality. Experimental results demonstrate that our QHD-based approach achieves superior modularity scores (up to 5.49\%) improvement with reduced computational overhead compared to traditional optimization methods. This work establishes QHD as an effective quantum-inspired framework for tackling graph partition challenges in large-scale networks.
Authors: Shuming Liang, Yu Ding, Zhidong Li, Bin Liang, Siqi Zhang, Yang Wang, Fang Chen
Abstract: This paper explores the ability of Graph Neural Networks (GNNs) in learning various forms of information for link prediction, alongside a brief review of existing link prediction methods. Our analysis reveals that GNNs cannot effectively learn structural information related to the number of common neighbors between two nodes, primarily due to the nature of set-based pooling of the neighborhood aggregation scheme. Also, our extensive experiments indicate that trainable node embeddings can improve the performance of GNN-based link prediction models. Importantly, we observe that the denser the graph, the greater such the improvement. We attribute this to the characteristics of node embeddings, where the link state of each link sample could be encoded into the embeddings of nodes that are involved in the neighborhood aggregation of the two nodes in that link sample. In denser graphs, every node could have more opportunities to attend the neighborhood aggregation of other nodes and encode states of more link samples to its embedding, thus learning better node embeddings for link prediction. Lastly, we demonstrate that the insights gained from our research carry important implications in identifying the limitations of existing link prediction methods, which could guide the future development of more robust algorithms.
Authors: Haiming Zhang, Wending Zhou, Yiyao Zhu, Xu Yan, Jiantao Gao, Dongfeng Bai, Yingjie Cai, Bingbing Liu, Shuguang Cui, Zhen Li
Abstract: This paper introduces VisionPAD, a novel self-supervised pre-training paradigm designed for vision-centric algorithms in autonomous driving. In contrast to previous approaches that employ neural rendering with explicit depth supervision, VisionPAD utilizes more efficient 3D Gaussian Splatting to reconstruct multi-view representations using only images as supervision. Specifically, we introduce a self-supervised method for voxel velocity estimation. By warping voxels to adjacent frames and supervising the rendered outputs, the model effectively learns motion cues in the sequential data. Furthermore, we adopt a multi-frame photometric consistency approach to enhance geometric perception. It projects adjacent frames to the current frame based on rendered depths and relative poses, boosting the 3D geometric representation through pure image supervision. Extensive experiments on autonomous driving datasets demonstrate that VisionPAD significantly improves performance in 3D object detection, occupancy prediction and map segmentation, surpassing state-of-the-art pre-training strategies by a considerable margin.
Authors: Jiatong Li, Yunqing Liu, Wei Liu, Jingdi Le, Di Zhang, Wenqi Fan, Dongzhan Zhou, Yuqiang Li, Qing Li
Abstract: Molecule discovery is a pivotal research field, impacting everything from the medicines we take to the materials we use. Recently, Large Language Models (LLMs) have been widely adopted in molecule understanding and generation, yet the alignments between molecules and their corresponding captions remain a significant challenge. Previous endeavours often treat the molecule as a general SMILES string or molecular graph, neglecting the fine-grained alignments between the molecular sub-structures and the descriptive textual phrases, which are crucial for accurate and explainable predictions. In this case, we introduce MolReFlect, a novel teacher-student framework designed to contextually perform the molecule-caption alignments in a fine-grained way. Our approach initially leverages a larger teacher LLM to label the detailed alignments by directly extracting critical phrases from molecule captions or SMILES strings and implying them to corresponding sub-structures or characteristics. To refine these alignments, we propose In-Context Selective Reflection, which retrieves previous extraction results as context examples for teacher LLM to reflect and lets a smaller student LLM select from in-context reflection and previous extraction results. Finally, we enhance the learning process of the student LLM through Chain-of-Thought In-Context Molecule Tuning, integrating the fine-grained alignments and the reasoning processes within the Chain-of-Thought format. Our experimental results demonstrate that MolReFlect enables LLMs like Mistral-7B to significantly outperform the previous baselines, achieving SOTA performance on the ChEBI-20 dataset. This advancement not only enhances the generative capabilities of LLMs in the molecule-caption translation task, but also contributes to a more explainable framework.
Authors: Feng Chen, Chenhui Gou, Jing Liu, Yang Yang, Zhaoyang Li, Jiyuan Zhang, Zhenbang Sun, Bohan Zhuang, Qi Wu
Abstract: As multimodal large language models (MLLMs) advance rapidly, rigorous evaluation has become essential, providing further guidance for their development. In this work, we focus on a unified and robust evaluation of \textbf{vision perception} abilities, the foundational skill of MLLMs. We find that existing perception benchmarks, each focusing on different question types, domains, and evaluation metrics, introduce significant evaluation variance, complicating comprehensive assessments of perception abilities when relying on any single benchmark. To address this, we introduce \textbf{AbilityLens}, a unified benchmark designed to evaluate MLLMs across six key perception abilities, focusing on both accuracy and stability, with each ability encompassing diverse question types, domains, and metrics. With the assistance of AbilityLens, we: (1) identify the strengths and weaknesses of current models, highlighting stability patterns and revealing a notable performance gap between open-source and closed-source models; (2) introduce an online evaluation mode, which uncovers interesting ability conflict and early convergence phenomena during MLLM training; and (3) design a simple ability-specific model merging method that combines the best ability checkpoint from early training stages, effectively mitigating performance decline due to ability conflict. The benchmark and online leaderboard will be released soon.
Authors: Xiaomin Li, Junyi Sha
Abstract: Identifying key product features that influence consumer preferences is essential in the fashion industry. In this study, we introduce a robust methodology to ascertain the most impactful features in fashion product images, utilizing past market sales data. First, we propose the metric called "influence score" to quantitatively assess the importance of product features. Then we develop a forecasting model, the Fashion Demand Predictor (FDP), which integrates Transformer-based models and Random Forest to predict market popularity based on product images. We employ image-editing diffusion models to modify these images and perform an ablation study, which validates the impact of the highest and lowest-scoring features on the model's popularity predictions. Additionally, we further validate these results through surveys that gather human rankings of preferences, confirming the accuracy of the FDP model's predictions and the efficacy of our method in identifying influential features. Notably, products enhanced with "good" features show marked improvements in predicted popularity over their modified counterparts. Our approach develops a fully automated and systematic framework for fashion image analysis that provides valuable guidance for downstream tasks such as fashion product design and marketing strategy development.
Authors: Yi Wang, Jiaze Wang, Ziyu Guo, Renrui Zhang, Donghao Zhou, Guangyong Chen, Anfeng Liu, Pheng-Ann Heng
Abstract: Recently Transformer-based models have advanced point cloud understanding by leveraging self-attention mechanisms, however, these methods often overlook latent information in less prominent regions, leading to increased sensitivity to perturbations and limited global comprehension. To solve this issue, we introduce PointACL, an attention-driven contrastive learning framework designed to address these limitations. Our method employs an attention-driven dynamic masking strategy that guides the model to focus on under-attended regions, enhancing the understanding of global structures within the point cloud. Then we combine the original pre-training loss with a contrastive learning loss, improving feature discrimination and generalization. Extensive experiments validate the effectiveness of PointACL, as it achieves state-of-the-art performance across a variety of 3D understanding tasks, including object classification, part segmentation, and few-shot learning. Specifically, when integrated with different Transformer backbones like Point-MAE and PointGPT, PointACL demonstrates improved performance on datasets such as ScanObjectNN, ModelNet40, and ShapeNetPart. This highlights its superior capability in capturing both global and local features, as well as its enhanced robustness against perturbations and incomplete data.
Authors: He Jia
Abstract: A major challenge in extracting information from current and upcoming surveys of cosmological Large-Scale Structure (LSS) is the limited availability of computationally expensive high-fidelity simulations. We introduce Neural Quantile Estimation (NQE), a new Simulation-Based Inference (SBI) method that leverages a large number of approximate simulations for training and a small number of high-fidelity simulations for calibration. This approach guarantees an unbiased posterior and achieves near-optimal constraining power when the approximate simulations are reasonably accurate. As a proof of concept, we demonstrate that cosmological parameters can be inferred at field level from projected 2-dim dark matter density maps up to $k_{\rm max}\sim1.5\,h$/Mpc at $z=0$ by training on $\sim10^4$ Particle-Mesh (PM) simulations with transfer function correction and calibrating with $\sim10^2$ Particle-Particle (PP) simulations. The calibrated posteriors closely match those obtained by directly training on $\sim10^4$ expensive PP simulations, but at a fraction of the computational cost. Our method offers a practical and scalable framework for SBI of cosmological LSS, enabling precise inference across vast volumes and down to small scales.
Authors: Kaito Shiku, Kazuya Nishimura, Daiki Suehiro, Kiyohito Tanaka, Ryoma Bise
Abstract: Patient-level diagnosis of severity in ulcerative colitis (UC) is common in real clinical settings, where the most severe score in a patient is recorded. However, previous UC classification methods (i.e., image-level estimation) mainly assumed the input was a single image. Thus, these methods can not utilize severity labels recorded in real clinical settings. In this paper, we propose a patient-level severity estimation method by a transformer with selective aggregator tokens, where a severity label is estimated from multiple images taken from a patient, similar to a clinical setting. Our method can effectively aggregate features of severe parts from a set of images captured in each patient, and it facilitates improving the discriminative ability between adjacent severity classes. Experiments demonstrate the effectiveness of the proposed method on two datasets compared with the state-of-the-art MIL methods. Moreover, we evaluated our method in real clinical settings and confirmed that our method outperformed the previous image-level methods. The code is publicly available at https://github.com/Shiku-Kaito/Ordinal-Multiple-instance-Learning-for-Ulcerative-Colitis-Severity-Estimation.
Authors: Sen Yang, Minyue Jiang, Ziwei Fan, Xiaolu Xie, Xiao Tan, Yingying Li, Errui Ding, Liang Wang, Jingdong Wang
Abstract: Recent advances in autonomous driving systems have shifted towards reducing reliance on high-definition maps (HDMaps) due to the huge costs of annotation and maintenance. Instead, researchers are focusing on online vectorized HDMap construction using on-board sensors. However, sensor-only approaches still face challenges in long-range perception due to the restricted views imposed by the mounting angles of onboard cameras, just as human drivers also rely on bird's-eye-view navigation maps for a comprehensive understanding of road structures. To address these issues, we propose to train the perception model to "see" standard definition maps (SDMaps). We encode SDMap elements into neural spatial map representations and instance tokens, and then incorporate such complementary features as prior information to improve the bird's eye view (BEV) feature for lane geometry and topology decoding. Based on the lane segment representation framework, the model simultaneously predicts lanes, centrelines and their topology. To further enhance the ability of geometry prediction and topology reasoning, we also use a topology-guided decoder to refine the predictions by exploiting the mutual relationships between topological and geometric features. We perform extensive experiments on OpenLane-V2 datasets to validate the proposed method. The results show that our model outperforms state-of-the-art methods by a large margin, with gains of +6.7 and +9.1 on the mAP and topology metrics. Our analysis also reveals that models trained with SDMap noise augmentation exhibit enhanced robustness.
Authors: Nikoo Moradi, Andr\'e Ferreira, Behrus Puladi, Jens Kleesiek, Emad Fatemizadeh, Gijs Luijten, Victor Alves, Jan Egger
Abstract: Radiation therapy (RT) is essential in treating head and neck cancer (HNC), with magnetic resonance imaging(MRI)-guided RT offering superior soft tissue contrast and functional imaging. However, manual tumor segmentation is time-consuming and complex, and therfore remains a challenge. In this study, we present our solution as team TUMOR to the HNTS-MRG24 MICCAI Challenge which is focused on automated segmentation of primary gross tumor volumes (GTVp) and metastatic lymph node gross tumor volume (GTVn) in pre-RT and mid-RT MRI images. We utilized the HNTS-MRG2024 dataset, which consists of 150 MRI scans from patients diagnosed with HNC, including original and registered pre-RT and mid-RT T2-weighted images with corresponding segmentation masks for GTVp and GTVn. We employed two state-of-the-art models in deep learning, nnUNet and MedNeXt. For Task 1, we pretrained models on pre-RT registered and mid-RT images, followed by fine-tuning on original pre-RT images. For Task 2, we combined registered pre-RT images, registered pre-RT segmentation masks, and mid-RT data as a multi-channel input for training. Our solution for Task 1 achieved 1st place in the final test phase with an aggregated Dice Similarity Coefficient of 0.8254, and our solution for Task 2 ranked 8th with a score of 0.7005. The proposed solution is publicly available at Github Repository.
Authors: Huiwon Jang, Sihyun Yu, Jinwoo Shin, Pieter Abbeel, Younggyo Seo
Abstract: Efficient tokenization of videos remains a challenge in training vision models that can process long videos. One promising direction is to develop a tokenizer that can encode long video clips, as it would enable the tokenizer to leverage the temporal coherence of videos better for tokenization. However, training existing tokenizers on long videos often incurs a huge training cost as they are trained to reconstruct all the frames at once. In this paper, we introduce CoordTok, a video tokenizer that learns a mapping from coordinate-based representations to the corresponding patches of input videos, inspired by recent advances in 3D generative models. In particular, CoordTok encodes a video into factorized triplane representations and reconstructs patches that correspond to randomly sampled $(x,y,t)$ coordinates. This allows for training large tokenizer models directly on long videos without requiring excessive training resources. Our experiments show that CoordTok can drastically reduce the number of tokens for encoding long video clips. For instance, CoordTok can encode a 128-frame video with 128$\times$128 resolution into 1280 tokens, while baselines need 6144 or 8192 tokens to achieve similar reconstruction quality. We further show that this efficient video tokenization enables memory-efficient training of a diffusion transformer that can generate 128 frames at once.
Authors: Youwei Zhou, Tianyang Xu, Cong Wu, Xiaojun Wu, Josef Kittler
Abstract: The shared topology of human skeletons motivated the recent investigation of graph convolutional network (GCN) solutions for action recognition. However, the existing GCNs rely on the binary connection of two neighbouring vertices (joints) formed by an edge (bone), overlooking the potential of constructing multi-vertex convolution structures. In this paper we address this oversight and explore the merits of a hyper-graph convolutional network (Hyper-GCN) to achieve the aggregation of rich semantic information conveyed by skeleton vertices. In particular, our Hyper-GCN adaptively optimises multi-scale hyper-graphs during training, revealing the action-driven multi-vertex relations. Besides, virtual connections are often designed to support efficient feature aggregation, implicitly extending the spectrum of dependencies within the skeleton. By injecting virtual connections into hyper-graphs, the semantic clues of diverse action categories can be highlighted. The results of experiments conducted on the NTU-60, NTU-120, and NW-UCLA datasets, demonstrate the merits of our Hyper-GCN, compared to the state-of-the-art methods. Specifically, we outperform the existing solutions on NTU-120, achieving 90.2\% and 91.4\% in terms of the top-1 recognition accuracy on X-Sub and X-Set.
Authors: Shulin Lan, Kanlin Liu, Yazhou Zhao, Chen Yang, Yingchao Wang, Xingshan Yao, Liehuang Zhu
Abstract: Current passive deepfake face-swapping detection methods encounter significance bottlenecks in model generalization capabilities. Meanwhile, proactive detection methods often use fixed watermarks which lack a close relationship with the content they protect and are vulnerable to security risks. Dynamic watermarks based on facial features offer a promising solution, as these features provide unique identifiers. Therefore, this paper proposes a Facial Feature-based Proactive deepfake detection method (FaceProtect), which utilizes changes in facial characteristics during deepfake manipulation as a novel detection mechanism. We introduce a GAN-based One-way Dynamic Watermark Generating Mechanism (GODWGM) that uses 128-dimensional facial feature vectors as inputs. This method creates irreversible mappings from facial features to watermarks, enhancing protection against various reverse inference attacks. Additionally, we propose a Watermark-based Verification Strategy (WVS) that combines steganography with GODWGM, allowing simultaneous transmission of the benchmark watermark representing facial features within the image. Experimental results demonstrate that our proposed method maintains exceptional detection performance and exhibits high practicality on images altered by various deepfake techniques.
Authors: Luca Parolari, Elena Izzo, Lamberto Ballan
Abstract: Referring Expression Comprehension (REC) aims to identify a particular object in a scene by a natural language expression, and is an important topic in visual language understanding. State-of-the-art methods for this task are based on deep learning, which generally requires expensive and manually labeled annotations. Some works tackle the problem with limited-supervision learning or relying on Large Vision and Language Models. However, the development of techniques to synthesize labeled data is overlooked. In this paper, we propose a novel framework that generates artificial data for the REC task, taking into account both textual and visual modalities. At first, our pipeline processes existing data to create variations in the annotations. Then, it generates an image using altered annotations as guidance. The result of this pipeline is a new dataset, called Harlequin, made by more than 1M queries. This approach eliminates manual data collection and annotation, enabling scalability and facilitating arbitrary complexity. We pre-train three REC models on Harlequin, then fine-tuned and evaluated on human-annotated datasets. Our experiments show that the pre-training on artificial data is beneficial for performance.
Authors: Dengsheng Chen, Jie Hu, Tiezhu Yue, Xiaoming Wei
Abstract: Denoising with a Joint-Embedding Predictive Architecture (D-JEPA), an autoregressive model, has demonstrated outstanding performance in class-conditional image generation. However, the application of next-token prediction in high-resolution text-to-image generation remains underexplored. In this paper, we introduce D-JEPA$\cdot$T2I, an extension of D-JEPA incorporating flow matching loss, designed to enable data-efficient continuous resolution learning. D-JEPA$\cdot$T2I leverages a multimodal visual transformer to effectively integrate textual and visual features and adopts Visual Rotary Positional Embedding (VoPE) to facilitate continuous resolution learning. Furthermore, we devise a data feedback mechanism that significantly enhances data utilization efficiency. For the first time, we achieve state-of-the-art \textbf{high-resolution} image synthesis via next-token prediction. The experimental code and pretrained models will be open-sourced at \url{https://d-jepa.github.io/t2i}.
Authors: Yuhang Song, Mario Gianni, Chenguang Yang, Kunyang Lin, Te-Chuan Chiu, Anh Nguyen, Chun-Yi Lee
Abstract: This paper addresses the challenge of fine-grained alignment in Vision-and-Language Navigation (VLN) tasks, where robots navigate realistic 3D environments based on natural language instructions. Current approaches use contrastive learning to align language with visual trajectory sequences. Nevertheless, they encounter difficulties with fine-grained vision negatives. To enhance cross-modal embeddings, we introduce a novel Bayesian Optimization-based adversarial optimization framework for creating fine-grained contrastive vision samples. To validate the proposed methodology, we conduct a series of experiments to assess the effectiveness of the enriched embeddings on fine-grained vision negatives. We conduct experiments on two common VLN benchmarks R2R and REVERIE, experiments on the them demonstrate that these embeddings benefit navigation, and can lead to a promising performance enhancement. Our source code and trained models are available at: https://anonymous.4open.science/r/FGVLN.
Authors: Chenfan Qu, Yiwu Zhong, Fengjun Guo, Lianwen Jin
Abstract: Image manipulation can lead to misinterpretation of visual content, posing significant risks to information security. Image Manipulation Localization (IML) has thus received increasing attention. However, existing IML methods rely heavily on task-specific designs, making them perform well only on one target image type but are mostly random guessing on other image types, and even joint training on multiple image types causes significant performance degradation. This hinders the deployment for real applications as it notably increases maintenance costs and the misclassification of image types leads to serious error accumulation. To this end, we propose Omni-IML, the first generalist model to unify diverse IML tasks. Specifically, Omni-IML achieves generalism by adopting the Modal Gate Encoder and the Dynamic Weight Decoder to adaptively determine the optimal encoding modality and the optimal decoder filters for each sample. We additionally propose an Anomaly Enhancement module that enhances the features of tampered regions with box supervision and helps the generalist model to extract common features across different IML tasks. We validate our approach on IML tasks across three major scenarios: natural images, document images, and face images. Without bells and whistles, our Omni-IML achieves state-of-the-art performance on all three tasks with a single unified model, providing valuable strategies and insights for real-world application and future research in generalist image forensics. Our code will be publicly available.
Authors: Ana\"is Halin, S\'ebastien Pi\'erard, Renaud Vandeghen, Beno\^it G\'erin, Maxime Zanella, Martin Colot, Jan Held, Anthony Cioppa, Emmanuel Jean, Gianluca Bontempi, Sa\"id Mahmoudi, Beno\^it Macq, Marc Van Droogenbroeck
Abstract: Characterizing domains is essential for models analyzing dynamic environments, as it allows them to adapt to evolving conditions or to hand the task over to backup systems when facing conditions outside their operational domain. Existing solutions typically characterize a domain by solving a regression or classification problem, which limits their applicability as they only provide a limited summarized description of the domain. In this paper, we present a novel approach to domain characterization by characterizing domains as probability distributions. Particularly, we develop a method to predict the likelihood of different weather conditions from images captured by vehicle-mounted cameras by estimating distributions of physical parameters using normalizing flows. To validate our proposed approach, we conduct experiments within the context of autonomous vehicles, focusing on predicting the distribution of weather parameters to characterize the operational domain. This domain is characterized by physical parameters (absolute characterization) and arbitrarily predefined domains (relative characterization). Finally, we evaluate whether a system can safely operate in a target domain by comparing it to multiple source domains where safety has already been established. This approach holds significant potential, as accurate weather prediction and effective domain adaptation are crucial for autonomous systems to adjust to dynamic environmental conditions.
Authors: Camilo Chac\'on Sartori, Christian Blum, Filippo Bistaffa
Abstract: The fast advancement of Large Vision-Language Models (LVLMs) has shown immense potential. These models are increasingly capable of tackling abstract visual tasks. Geometric structures, particularly graphs with their inherent flexibility and complexity, serve as an excellent benchmark for evaluating these models' predictive capabilities. While human observers can readily identify subtle visual details and perform accurate analyses, our investigation reveals that state-of-the-art LVLMs exhibit consistent limitations in specific visual graph scenarios, especially when confronted with stylistic variations. In response to these challenges, we introduce VisGraphVar (Visual Graph Variability), a customizable benchmark generator able to produce graph images for seven distinct task categories (detection, classification, segmentation, pattern recognition, link prediction, reasoning, matching), designed to systematically evaluate the strengths and limitations of individual LVLMs. We use VisGraphVar to produce 990 graph images and evaluate six LVLMs, employing two distinct prompting strategies, namely zero-shot and chain-of-thought. The findings demonstrate that variations in visual attributes of images (e.g., node labeling and layout) and the deliberate inclusion of visual imperfections, such as overlapping nodes, significantly affect model performance. This research emphasizes the importance of a comprehensive evaluation across graph-related tasks, extending beyond reasoning alone. VisGraphVar offers valuable insights to guide the development of more reliable and robust systems capable of performing advanced visual graph analysis.
Authors: Giorgio Palma, Andrea Serani, Kevin McTaggart, Shawn Aram, David W. Wundrow, David Drazen, Matteo Diez
Abstract: Digital twins are widely considered enablers of groundbreaking changes in the development, operation, and maintenance of novel generations of products. They are meant to provide reliable and timely predictions to inform decisions along the entire product life cycle. One of their most interesting applications in the naval field is the digital twinning of ship performances in waves, a crucial aspect in design and operation safety. In this paper, a Bayesian extension of the Hankel dynamic mode decomposition method is proposed for ship motion's nowcasting as a prediction tool for naval digital twins. The proposed algorithm meets all the requirements for formulations devoted to digital twinning, being able to adapt the resulting models with the data incoming from the physical system, using a limited amount of data, producing real-time predictions, and estimating their reliability. Results are presented and discussed for the course-keeping of the 5415M model in beam-quartering sea state 7 irregular waves at Fr = 0.33, using data from three different CFD solvers. The results show predictions keeping good accuracy levels up to five wave encounter periods, with the Bayesian formulation improving the deterministic forecasts. In addition, a connection between the predicted uncertainty and prediction accuracy is found.
Authors: Jeongsol Kim, Beomsu Kim, Jong Chul Ye
Abstract: Diffusion models (DMs), which enable both image generation from noise and inversion from data, have inspired powerful unpaired image-to-image (I2I) translation algorithms. However, they often require a larger number of neural function evaluations (NFEs), limiting their practical applicability. In this paper, we tackle this problem with Schrodinger Bridges (SBs), which are stochastic differential equations (SDEs) between distributions with minimal transport cost. We analyze the probability flow ordinary differential equation (ODE) formulation of SBs, and observe that we can decompose its vector field into a linear combination of source predictor, target predictor, and noise predictor. Inspired by this observation, we propose Latent Schrodinger Bridges (LSBs) that approximate the SB ODE via pre-trained Stable Diffusion, and develop appropriate prompt optimization and change of variables formula to match the training and inference between distributions. We demonstrate that our algorithm successfully conduct competitive I2I translation in unsupervised setting with only a fraction of computation cost required by previous DM-based I2I methods.
Authors: Xuewu Lin, Tianwei Lin, Lichao Huang, Hongyu Xie, Zhizhong Su
Abstract: In embodied intelligence systems, a key component is 3D perception algorithm, which enables agents to understand their surrounding environments. Previous algorithms primarily rely on point cloud, which, despite offering precise geometric information, still constrain perception performance due to inherent sparsity, noise, and data scarcity. In this work, we introduce a novel image-centric 3D perception model, BIP3D, which leverages expressive image features with explicit 3D position encoding to overcome the limitations of point-centric methods. Specifically, we leverage pre-trained 2D vision foundation models to enhance semantic understanding, and introduce a spatial enhancer module to improve spatial understanding. Together, these modules enable BIP3D to achieve multi-view, multi-modal feature fusion and end-to-end 3D perception. In our experiments, BIP3D outperforms current state-of-the-art results on the EmbodiedScan benchmark, achieving improvements of 5.69% in the 3D detection task and 15.25% in the 3D visual grounding task.
Authors: Sebastian Stock, Jannik Dunkelau, Atif Mashkoor
Abstract: With artificial intelligence (AI) being well established within the daily lives of research communities, we turn our gaze toward an application area that appears intuitively unsuited for probabilistic decision-making: the area of formal methods (FM). FM aim to provide sound and understandable reasoning about problems in computer science, which seemingly collides with the black-box nature that inhibits many AI approaches. However, many researchers have crossed this gap and applied AI techniques to enhance FM approaches. As this dichotomy of FM and AI sparked our interest, we conducted a systematic mapping study to map the current landscape of research publications. In this study, we investigate the previous five years of applied AI to FM (2019-2023), as these correspond to periods of high activity. This investigation results in 189 entries, which we explore in more detail to find current trends, highlight research gaps, and give suggestions for future research.
Authors: Dingyuan Shi, Yong Wang, Hangyu Li, Xiangxiang Chu
Abstract: Diffusion models have shown remarkable success in text-to-image generation, making alignment methods for these models increasingly important. A key challenge is the sparsity of preference labels, which are typically available only at the terminal of denoising trajectories. This raises the issue of how to assign credit across denoising steps based on these sparse labels. In this paper, we propose Denoised Distribution Estimation (DDE), a novel method for credit assignment. Unlike previous approaches that rely on auxiliary models or hand-crafted schemes, DDE derives its strategy more explicitly. The proposed DDE directly estimates the terminal denoised distribution from the perspective of each step. It is equipped with two estimation strategies and capable of representing the entire denoising trajectory with a single model inference. Theoretically and empirically, we show that DDE prioritizes optimizing the middle part of the denoising trajectory, resulting in a novel and effective credit assignment scheme. Extensive experiments demonstrate that our approach achieves superior performance, both quantitatively and qualitatively.
Authors: Yanyun Ding, Zhenghua Yao, Peili Li, Yunhai Xiao
Abstract: The elastic net penalty is frequently employed in high-dimensional statistics for parameter regression and variable selection. It is particularly beneficial compared to lasso when the number of predictors greatly surpasses the number of observations. However, empirical evidence has shown that the $\ell_q$-norm penalty (where $0 < q < 1$) often provides better regression compared to the $\ell_1$-norm penalty, demonstrating enhanced robustness in various scenarios. In this paper, we explore a generalized elastic net model that employs a $\ell_r$-norm (where $r \geq 1$) in loss function to accommodate various types of noise, and employs a $\ell_q$-norm (where $0 < q < 1$) to replace the $\ell_1$-norm in elastic net penalty. Theoretically, we establish the computable lower bounds for the nonzero entries of the generalized first-order stationary points of the proposed generalized elastic net model. For implementation, we develop two efficient algorithms based on the locally Lipschitz continuous $\epsilon$-approximation to $\ell_q$-norm. The first algorithm employs an alternating direction method of multipliers (ADMM), while the second utilizes a proximal majorization-minimization method (PMM), where the subproblems are addressed using the semismooth Newton method (SNN). We also perform extensive numerical experiments with both simulated and real data, showing that both algorithms demonstrate superior performance. Notably, the PMM-SSN is efficient than ADMM, even though the latter provides a simpler implementation.
Authors: Yuheng Xu, Taiping Zhang
Abstract: Domain-invariant representation learning is a powerful method for domain generalization. Previous approaches face challenges such as high computational demands, training instability, and limited effectiveness with high-dimensional data, potentially leading to the loss of valuable features. To address these issues, we hypothesize that an ideal generalized representation should exhibit similar pattern responses within the same channel across cross-domain images. Based on this hypothesis, we use deep features from the source domain as queries, and deep features from the generated domain as keys and values. Through a cross-channel attention mechanism, the original deep features are reconstructed into robust regularization representations, forming an explicit constraint that guides the model to learn domain-invariant representations. Additionally, style augmentation is another common method. However, existing methods typically generate new styles through convex combinations of source domains, which limits the diversity of training samples by confining the generated styles to the original distribution. To overcome this limitation, we propose an Adaptive Feature Blending (AFB) method that generates out-of-distribution samples while exploring the in-distribution space, significantly expanding the domain range. Extensive experimental results demonstrate that our proposed methods achieve superior performance on two standard domain generalization benchmarks for medical image segmentation.
Authors: Juan Miguel Lopez Alcaraz, Nils Strodthoff
Abstract: Background: Laboratory values are fundamental to medical diagnosis and management, but acquiring these values can be costly, invasive, and time-consuming. While electrocardiogram (ECG) patterns have been linked to certain laboratory abnormalities, the comprehensive modeling of these relationships remains underexplored. Methods: We utilize MIMIC-IV dataset to develop multimodal deep-learning models to demonstrate the feasibility of estimating (real-time) and monitoring (predict at future intervals) laboratory value abnormalities from ECG waveforms, demographics, biometrics, and vital signs. Results: The models exhibit a strong predictive performance with AUROC scores above 0.70 in a statistically significant manner for 23 laboratory values in the estimation setting and up to 26 values in the monitoring setting. Most notably, the accurately predictable values encompassing abnormalities across diverse physiological categories such as cardiac, renal, hematological, metabolic, immunological and coagulation. To name examples, for estimation NTproBNP (>353 pg/mL) with 0.882, whereas for monitoring at 30 minutes Urea nitrogen (<6 mg/dL) with 0.851, at 60 minutes creatinine (<0.5 mg/dL) with 0.85, and at 120 minutes hemoglobin (>17.5 g/dL) with 0.821. Conclusions: This study provides first evidence for the feasibility of using ECG data alongside clinical routine data for the real-time estimation and monitoring of laboratory value abnormalities, which could provide a non-invasive, cost-effective supplement to traditional laboratory testing, with strong implications for enhanced patient monitoring and early intervention. Further validation could facilitate their integration into routine clinical practice.
Authors: Zhenwei Yang, Jilei Mao, Wenxian Yang, Yibo Ai, Yu Kong, Haibao Yu, Weidong Zhang
Abstract: Temporal perception, the ability to detect and track objects over time, is critical in autonomous driving for maintaining a comprehensive understanding of dynamic environments. However, this task is hindered by significant challenges, including incomplete perception caused by occluded objects and observational blind spots, which are common in single-vehicle perception systems. To address these issues, we introduce LET-VIC, a LiDAR-based End-to-End Tracking framework for Vehicle-Infrastructure Cooperation (VIC). LET-VIC leverages Vehicle-to-Everything (V2X) communication to enhance temporal perception by fusing spatial and temporal data from both vehicle and infrastructure sensors. First, it spatially integrates Bird's Eye View (BEV) features from vehicle-side and infrastructure-side LiDAR data, creating a comprehensive view that mitigates occlusions and compensates for blind spots. Second, LET-VIC incorporates temporal context across frames, allowing the model to leverage historical data for enhanced tracking stability and accuracy. To further improve robustness, LET-VIC includes a Calibration Error Compensation (CEC) module to address sensor misalignments and ensure precise feature alignment. Experiments on the V2X-Seq-SPD dataset demonstrate that LET-VIC significantly outperforms baseline models, achieving at least a 13.7% improvement in mAP and a 13.1% improvement in AMOTA without considering communication delays. This work offers a practical solution and a new research direction for advancing temporal perception in autonomous driving through vehicle-infrastructure cooperation.
Authors: Koji Hashimoto, Koshiro Matsuo, Masaki Murata, Gakuto Ogiwara
Abstract: Topological solitons, which are stable, localized solutions of nonlinear differential equations, are crucial in various fields of physics and mathematics, including particle physics and cosmology. However, solving these solitons presents significant challenges due to the complexity of the underlying equations and the computational resources required for accurate solutions. To address this, we have developed a novel method using neural network (NN) to efficiently solve solitons. A similar NN approach is Physics-Informed Neural Networks (PINN). In a comparative analysis between our method and PINN, we find that our method achieves shorter computation times while maintaining the same level of accuracy. This advancement in computational efficiency not only overcomes current limitations but also opens new avenues for studying topological solitons and their dynamical behavior.
Authors: Lars Nieradzik, Henrike Stephani, Janis Keuper
Abstract: In this paper, we present an approach for evaluating attribution maps, which play a central role in interpreting the predictions of convolutional neural networks (CNNs). We show that the widely used insertion/deletion metrics are susceptible to distribution shifts that affect the reliability of the ranking. Our method proposes to replace pixel modifications with adversarial perturbations, which provides a more robust evaluation framework. By using smoothness and monotonicity measures, we illustrate the effectiveness of our approach in correcting distribution shifts. In addition, we conduct the most comprehensive quantitative and qualitative assessment of attribution maps to date. Introducing baseline attribution maps as sanity checks, we find that our metric is the only contender to pass all checks. Using Kendall's $\tau$ rank correlation coefficient, we show the increased consistency of our metric across 15 dataset-architecture combinations. Of the 16 attribution maps tested, our results clearly show SmoothGrad to be the best map currently available. This research makes an important contribution to the development of attribution maps by providing a reliable and consistent evaluation framework. To ensure reproducibility, we will provide the code along with our results.
Authors: Miriam Alber, Christoph H\"ones, Patrick Baier
Abstract: One of the most promising use-cases for machine learning in industrial manufacturing is the early detection of defective products using a quality control system. Such a system can save costs and reduces human errors due to the monotonous nature of visual inspections. Today, a rich body of research exists which employs machine learning methods to identify rare defective products in unbalanced visual quality control datasets. These methods typically rely on two components: A visual backbone to capture the features of the input image and an anomaly detection algorithm that decides if these features are within an expected distribution. With the rise of transformer architecture as visual backbones of choice, there exists now a great variety of different combinations of these two components, ranging all along the trade-off between detection quality and inference time. Facing this variety, practitioners in the field often have to spend a considerable amount of time on researching the right combination for their use-case at hand. Our contribution is to help practitioners with this choice by reviewing and evaluating current vision transformer models together with anomaly detection methods. For this, we chose SotA models of both disciplines, combined them and evaluated them towards the goal of having small, fast and efficient anomaly detection models suitable for industrial manufacturing. We evaluated the results of our experiments on the well-known MVTecAD and BTAD datasets. Moreover, we give guidelines for choosing a suitable model architecture for a quality control system in practice, considering given use-case and hardware constraints.
Authors: Alec Wright, Alistair Carson, Lauri Juvela
Abstract: This paper introduces Open-Amp, a synthetic data framework for generating large-scale and diverse audio effects data. Audio effects are relevant to many musical audio processing and Music Information Retrieval (MIR) tasks, such as modelling of analog audio effects, automatic mixing, tone matching and transcription. Existing audio effects datasets are limited in scope, usually including relatively few audio effects processors and a limited amount of input audio signals. Our proposed framework overcomes these issues, by crowdsourcing neural network emulations of guitar amplifiers and effects, created by users of open-source audio effects emulation software. This allows users of Open-Amp complete control over the input signals to be processed by the effects models, as well as providing high-quality emulations of hundreds of devices. Open-Amp can render audio online during training, allowing great flexibility in data augmentation. Our experiments show that using Open-Amp to train a guitar effects encoder achieves new state-of-the-art results on multiple guitar effects classification tasks. Furthermore, we train a one-to-many guitar effects model using Open-Amp, and use it to emulate unseen analog effects via manipulation of its learned latent space, indicating transferability to analog guitar effects data.
Authors: Jos\'ephine Pazem, Marius Krumm, Alexander Q. Vining, Lukas J. Fiderer, Hans J. Briegel
Abstract: In the last decade, the free energy principle (FEP) and active inference (AIF) have achieved many successes connecting conceptual models of learning and cognition to mathematical models of perception and action. This effort is driven by a multidisciplinary interest in understanding aspects of self-organizing complex adaptive systems, including elements of agency. Various reinforcement learning (RL) models performing active inference have been proposed and trained on standard RL tasks using deep neural networks. Recent work has focused on improving such agents' performance in complex environments by incorporating the latest machine learning techniques. In this paper, we take an alternative approach. Within the constraints imposed by the FEP and AIF, we attempt to model agents in an interpretable way without deep neural networks by introducing Free Energy Projective Simulation (FEPS). Using internal rewards only, FEPS agents build a representation of their partially observable environments with which they interact. Following AIF, the policy to achieve a given task is derived from this world model by minimizing the expected free energy. Leveraging the interpretability of the model, techniques are introduced to deal with long-term goals and reduce prediction errors caused by erroneous hidden state estimation. We test the FEPS model on two RL environments inspired from behavioral biology: a timed response task and a navigation task in a partially observable grid. Our results show that FEPS agents fully resolve the ambiguity of both environments by appropriately contextualizing their observations based on prediction accuracy only. In addition, they infer optimal policies flexibly for any target observation in the environment.
Authors: Tsogt-Ochir Enkhbayar
Abstract: This paper advances the computational efficiency of Deep Hedging frameworks through the novel integration of Kronecker-Factored Approximate Curvature (K-FAC) optimization. While recent literature has established Deep Hedging as a data-driven alternative to traditional risk management strategies, the computational burden of training neural networks with first-order methods remains a significant impediment to practical implementation. The proposed architecture couples Long Short-Term Memory (LSTM) networks with K-FAC second-order optimization, specifically addressing the challenges of sequential financial data and curvature estimation in recurrent networks. Empirical validation using simulated paths from a calibrated Heston stochastic volatility model demonstrates that the K-FAC implementation achieves marked improvements in convergence dynamics and hedging efficacy. The methodology yields a 78.3% reduction in transaction costs ($t = 56.88$, $p < 0.001$) and a 34.4% decrease in profit and loss (P&L) variance compared to Adam optimization. Moreover, the K-FAC-enhanced model exhibits superior risk-adjusted performance with a Sharpe ratio of 0.0401, contrasting with $-0.0025$ for the baseline model. These results provide compelling evidence that second-order optimization methods can materially enhance the tractability of Deep Hedging implementations. The findings contribute to the growing literature on computational methods in quantitative finance while highlighting the potential for advanced optimization techniques to bridge the gap between theoretical frameworks and practical applications in financial markets.
Authors: Sneha Sudhir Shetiya, Divya Garikapati, Veeraja Sohoni
Abstract: Functional safety forms an important aspect in the design of systems. Its emphasis on the automotive industry has evolved significantly over the years. Till date many methods have been developed to get appropriate FTA(Fault Tree analysis) for various scenarios and features pertaining to Autonomous Driving. This paper is an attempt to explore the scope of using Generative Artificial Intelligence(GenAI) in order to develop Fault Tree Analysis(FTA) with the use case of malfunction for the Lidar sensor in mind. We explore various available open source Large Language Models(LLM) models and then dive deep into one of them to study its responses and provide our analysis. This paper successfully shows the possibility to train existing Large Language models through Prompt Engineering for fault tree analysis for any Autonomy usecase aided with PlantUML tool.
Authors: Keda Tao, Can Qin, Haoxuan You, Yang Sui, Huan Wang
Abstract: Video large language models (VLLMs) have significantly advanced recently in processing complex video content, yet their inference efficiency remains constrained because of the high computational cost stemming from the thousands of visual tokens generated from the video inputs. We empirically observe that, unlike single image inputs, VLLMs typically attend visual tokens from different frames at different decoding iterations, making a one-shot pruning strategy prone to removing important tokens by mistake. Motivated by this, we present DyCoke, a training-free token compression method to optimize token representation and accelerate VLLMs. DyCoke incorporates a plug-and-play temporal compression module to minimize temporal redundancy by merging redundant tokens across frames, and applies dynamic KV cache reduction to prune spatially redundant tokens selectively. It ensures high-quality inference by dynamically retaining the critical tokens at each decoding step. Extensive experimental results demonstrate that DyCoke can outperform the prior SoTA counterparts, achieving 1.5X inference speedup, 1.4X memory reduction against the baseline VLLM, while still improving the performance, with no training.
Authors: Yu Xu, Fan Tang, Juan Cao, Yuxin Zhang, Xiaoyu Kong, Jintao Li, Oliver Deussen, Tong-Yee Lee
Abstract: Diffusion Transformers (DiTs) have exhibited robust capabilities in image generation tasks. However, accurate text-guided image editing for multimodal DiTs (MM-DiTs) still poses a significant challenge. Unlike UNet-based structures that could utilize self/cross-attention maps for semantic editing, MM-DiTs inherently lack support for explicit and consistent incorporated text guidance, resulting in semantic misalignment between the edited results and texts. In this study, we disclose the sensitivity of different attention heads to different image semantics within MM-DiTs and introduce HeadRouter, a training-free image editing framework that edits the source image by adaptively routing the text guidance to different attention heads in MM-DiTs. Furthermore, we present a dual-token refinement module to refine text/image token representations for precise semantic guidance and accurate region expression. Experimental results on multiple benchmarks demonstrate HeadRouter's performance in terms of editing fidelity and image quality.
Authors: Valentin Barriere
Abstract: Deep Learning models tend to learn correlations of patterns on huge datasets. The bigger these systems are, the more complex are the phenomena they can detect, and the more data they need for this. The use of Artificial Intelligence (AI) is becoming increasingly ubiquitous in our society, and its impact is growing everyday. The promises it holds strongly depend on their fair and universal use, such as access to information or education for all. In a world of inequalities, they can help to reach the most disadvantaged areas. However, such a universal systems must be able to represent society, without benefiting some at the expense of others. We must not reproduce the inequalities observed throughout the world, but educate these IAs to go beyond them. We have seen cases where these systems use gender, race, or even class information in ways that are not appropriate for resolving their tasks. Instead of real causal reasoning, they rely on spurious correlations, which is what we usually call a bias. In this paper, we first attempt to define what is a bias in general terms. It helps us to demystify the concept of bias, to understand why we can find them everywhere and why they are sometimes useful. Second, we focus over the notion of what is generally seen as negative bias, the one we want to avoid in machine learning, before presenting a general zoology containing the most common of these biases. We finally conclude by looking at classical methods to detect them, by means of specially crafted datasets of templates and specific algorithms, and also classical methods to mitigate them.
Authors: Ji-Hun Oh, Kianoush Falahkheirkhah, Rohit Bhargava
Abstract: Significant biomedical research and clinical care rely on the histopathologic examination of tissue structure using microscopy of stained tissue. Virtual staining (VS) offers a promising alternative with the potential to reduce cost and eliminate the use of toxic reagents. However, the critical challenge of hallucinations limits confidence in its use, necessitating a VS co-pilot to detect these hallucinations. Here, we first formally establish the problem of hallucination detection in VS. Next, we introduce a scalable, post-hoc hallucination detection method that identifies a Neural Hallucination Precursor (NHP) from VS model embeddings for test-time detection. We report extensive validation across diverse and challenging VS settings to demonstrate NHP's effectiveness and robustness. Furthermore, we show that VS models with fewer hallucinations do not necessarily disclose them better, risking a false sense of security when reporting just the former metric. This highlights the need for a reassessment of current VS evaluation practices.
Authors: Zizhao Wu, Jian Shi, Xuan Deng, Cheng Zhang, Genfu Yang, Ming Zeng, Yunhai Wang
Abstract: Point cloud completion aims to infer a complete shape from its partial observation. Many approaches utilize a pure encoderdecoder paradigm in which complete shape can be directly predicted by shape priors learned from partial scans, however, these methods suffer from the loss of details inevitably due to the feature abstraction issues. In this paper, we propose a novel framework,termed SPAC-Net, that aims to rethink the completion task under the guidance of a new structural prior, we call it interface. Specifically, our method first investigates Marginal Detector (MAD) module to localize the interface, defined as the intersection between the known observation and the missing parts. Based on the interface, our method predicts the coarse shape by learning the displacement from the points in interface move to their corresponding position in missing parts. Furthermore, we devise an additional Structure Supplement(SSP) module before the upsampling stage to enhance the structural details of the coarse shape, enabling the upsampling module to focus more on the upsampling task. Extensive experiments have been conducted on several challenging benchmarks, and the results demonstrate that our method outperforms existing state-of-the-art approaches.
Authors: Razvan-Andrei Lascu, Mateusz B. Majka, David \v{S}i\v{s}ka, {\L}ukasz Szpruch
Abstract: We investigate proximal descent methods, inspired by the minimizing movement scheme introduced by Jordan, Kinderlehrer and Otto, for optimizing entropy-regularized functionals on the Wasserstein space. We establish linear convergence under flat convexity assumptions, thereby relaxing the common reliance on geodesic convexity. Our analysis circumvents the need for discrete-time adaptations of the Evolution Variational Inequality (EVI). Instead, we leverage a uniform logarithmic Sobolev inequality (LSI) and the entropy "sandwich" lemma, extending the analysis from arXiv:2201.10469 and arXiv:2202.01009. The major challenge in the proof via LSI is to show that the relative Fisher information $I(\cdot|\pi)$ is well-defined at every step of the scheme. Since the relative entropy is not Wasserstein differentiable, we prove that along the scheme the iterates belong to a certain class of Sobolev regularity, and hence the relative entropy $\operatorname{KL}(\cdot|\pi)$ has a unique Wasserstein sub-gradient, and that the relative Fisher information is indeed finite.
Authors: Jan Bednarik, Erroll Wood, Vasileios Choutas, Timo Bolkart, Daoye Wang, Chenglei Wu, Thabo Beeler
Abstract: Nowadays, it is possible to scan faces and automatically register them with high quality. However, the resulting face meshes often need further processing: we need to stabilize them to remove unwanted head movement. Stabilization is important for tasks like game development or movie making which require facial expressions to be cleanly separated from rigid head motion. Since manual stabilization is labor-intensive, there have been attempts to automate it. However, previous methods remain impractical: they either still require some manual input, produce imprecise alignments, rely on dubious heuristics and slow optimization, or assume a temporally ordered input. Instead, we present a new learning-based approach that is simple and fully automatic. We treat stabilization as a regression problem: given two face meshes, our network directly predicts the rigid transform between them that brings their skulls into alignment. We generate synthetic training data using a 3D Morphable Model (3DMM), exploiting the fact that 3DMM parameters separate skull motion from facial skin motion. Through extensive experiments we show that our approach outperforms the state-of-the-art both quantitatively and qualitatively on the tasks of stabilizing discrete sets of facial expressions as well as dynamic facial performances. Furthermore, we provide an ablation study detailing the design choices and best practices to help others adopt our approach for their own uses. Supplementary videos can be found on the project webpage syntec-research.github.io/FaceStab.
Authors: Irfan Nafiz Shahan, Pulok Ahmed Auvi
Abstract: Voice recognition and speaker identification are vital for applications in security and personal assistants. This paper presents a lightweight 1D-Convolutional Neural Network (1D-CNN) designed to perform speaker identification on minimal datasets. Our approach achieves a validation accuracy of 97.87%, leveraging data augmentation techniques to handle background noise and limited training samples. Future improvements include testing on larger datasets and integrating transfer learning methods to enhance generalizability. We provide all code, the custom dataset, and the trained models to facilitate reproducibility. These resources are available on our GitHub repository: https://github.com/IrfanNafiz/RecMe.
Authors: Lakshmikar R. Polamreddy, Kalyan Roy, Sheng-Han Yueh, Deepshikha Mahato, Shilpa Kuppili, Jialu Li, Youshan Zhang
Abstract: The scarcity of accessible medical image data poses a significant obstacle in effectively training deep learning models for medical diagnosis, as hospitals refrain from sharing their data due to privacy concerns. In response, we gathered a diverse dataset named MedImgs, which comprises over 250,127 images spanning 61 disease types and 159 classes of both humans and animals from open-source repositories. We propose a Leapfrog Latent Consistency Model (LLCM) that is distilled from a retrained diffusion model based on the collected MedImgs dataset, which enables our model to generate real-time high-resolution images. We formulate the reverse diffusion process as a probability flow ordinary differential equation (PF-ODE) and solve it in latent space using the Leapfrog algorithm. This formulation enables rapid sampling without necessitating additional iterations. Our model demonstrates state-of-the-art performance in generating medical images. Furthermore, our model can be fine-tuned with any custom medical image datasets, facilitating the generation of a vast array of images. Our experimental results outperform those of existing models on unseen dog cardiac X-ray images. Source code is available at https://github.com/lskdsjy/LeapfrogLCM.
Authors: E-Ro Nguyen, Hieu Le, Dimitris Samaras, Michael Ryoo
Abstract: Recent works on Generalized Referring Expression Segmentation (GRES) struggle with handling complex expressions referring to multiple distinct objects. This is because these methods typically employ an end-to-end foreground-background segmentation and lack a mechanism to explicitly differentiate and associate different object instances to the text query. To this end, we propose InstAlign, a method that incorporates object-level reasoning into the segmentation process. Our model leverages both text and image inputs to extract a set of object-level tokens that capture both the semantic information in the input prompt and the objects within the image. By modeling the text-object alignment via instance-level supervision, each token uniquely represents an object segment in the image, while also aligning with relevant semantic information from the text. Extensive experiments on the gRefCOCO and Ref-ZOM benchmarks demonstrate that our method significantly advances state-of-the-art performance, setting a new standard for precise and flexible GRES.
Authors: Robert A. Vandermeulen, Wai Ming Tai, Bryon Aragam
Abstract: We show that deep neural networks achieve dimension-independent rates of convergence for learning structured densities such as those arising in image, audio, video, and text applications. More precisely, we demonstrate that neural networks with a simple $L^2$-minimizing loss achieve a rate of $n^{-1/(4+r)}$ in nonparametric density estimation when the underlying density is Markov to a graph whose maximum clique size is at most $r$, and we provide evidence that in the aforementioned applications, this size is typically constant, i.e., $r=O(1)$. We then establish that the optimal rate in $L^1$ is $n^{-1/(2+r)}$ which, compared to the standard nonparametric rate of $n^{-1/(2+d)}$, reveals that the effective dimension of such problems is the size of the largest clique in the Markov random field. These rates are independent of the data's ambient dimension, making them applicable to realistic models of image, sound, video, and text data. Our results provide a novel justification for deep learning's ability to circumvent the curse of dimensionality, demonstrating dimension-independent convergence rates in these contexts.
Authors: Zhenxiong Tan, Songhua Liu, Xingyi Yang, Qiaochu Xue, Xinchao Wang
Abstract: In this paper, we introduce OminiControl, a highly versatile and parameter-efficient framework that integrates image conditions into pre-trained Diffusion Transformer (DiT) models. At its core, OminiControl leverages a parameter reuse mechanism, enabling the DiT to encode image conditions using itself as a powerful backbone and process them with its flexible multi-modal attention processors. Unlike existing methods, which rely heavily on additional encoder modules with complex architectures, OminiControl (1) effectively and efficiently incorporates injected image conditions with only ~0.1% additional parameters, and (2) addresses a wide range of image conditioning tasks in a unified manner, including subject-driven generation and spatially-aligned conditions such as edges, depth, and more. Remarkably, these capabilities are achieved by training on images generated by the DiT itself, which is particularly beneficial for subject-driven generation. Extensive evaluations demonstrate that OminiControl outperforms existing UNet-based and DiT-adapted models in both subject-driven and spatially-aligned conditional generation. Additionally, we release our training dataset, Subjects200K, a diverse collection of over 200,000 identity-consistent images, along with an efficient data synthesis pipeline to advance research in subject-consistent generation.
Authors: Karsten Roth, Zeynep Akata, Dima Damen, Ivana Bala\v{z}evi\'c, Olivier J. H\'enaff
Abstract: Large-scale multimodal representation learning successfully optimizes for zero-shot transfer at test time. Yet the standard pretraining paradigm (contrastive learning on large amounts of image-text data) does not explicitly encourage representations to support few-shot adaptation. In this work, we propose a simple, but carefully designed extension to multimodal pretraining which enables representations to accommodate additional context. Using this objective, we show that vision-language models can be trained to exhibit significantly increased few-shot adaptation: across 21 downstream tasks, we find up to four-fold improvements in test-time sample efficiency, and average few-shot adaptation gains of over 5%, while retaining zero-shot generalization performance across model scales and training durations. In particular, equipped with simple, training-free, metric-based adaptation mechanisms, our representations easily surpass more complex and expensive optimization-based schemes, vastly simplifying generalization to new domains.
Authors: Alexandros Stergiou, Ronald Poppe
Abstract: We have witnessed impressive advances in video action understanding. Increased dataset sizes, variability, and computation availability have enabled leaps in performance and task diversification. Current systems can provide coarse- and fine-grained descriptions of video scenes, extract segments corresponding to queries, synthesize unobserved parts of videos, and predict context. This survey comprehensively reviews advances in uni- and multi-modal action understanding across a range of tasks. We focus on prevalent challenges, overview widely adopted datasets, and survey seminal works with an emphasis on recent advances. We broadly distinguish between three temporal scopes: (1) recognition tasks of actions observed in full, (2) prediction tasks for ongoing partially observed actions, and (3) forecasting tasks for subsequent unobserved action. This division allows us to identify specific action modeling and video representation challenges. Finally, we outline future directions to address current shortcomings.
Authors: Afrah Fareaa, Mustafa Serdar Celebi
Abstract: We investigate the use of learnable activation functions in Physics-Informed Neural Networks (PINNs) for solving Partial Differential Equations (PDEs). Specifically, we compare the efficacy of traditional Multilayer Perceptrons (MLPs) with fixed and learnable activations against Kolmogorov-Arnold Networks (KANs), which employ learnable basis functions. Physics-informed neural networks (PINNs) have emerged as an effective method for directly incorporating physical laws into the learning process, offering a data-efficient solution for both the forward and inverse problems associated with PDEs. However, challenges such as effective training and spectral bias, where low-frequency components are learned more effectively, often limit their applicability to problems characterized by rapid oscillations or sharp transitions. By employing different activation or basis functions on MLP and KAN, we assess their impact on convergence behavior and spectral bias mitigation, and the accurate approximation of PDEs. The findings offer insights into the design of neural network architectures that balance training efficiency, convergence speed, and test accuracy for PDE solvers. By evaluating the influence of activation or basis function choices, this work provides guidelines for developing more robust and accurate PINN models. The source code and pre-trained models used in this study are made publicly available to facilitate reproducibility and future exploration.
Authors: Samarth N Ramesh, Zhixue Zhao
Abstract: As text-to-image models grow increasingly powerful and complex, their burgeoning size presents a significant obstacle to widespread adoption, especially on resource-constrained devices. This paper presents a pioneering study on post-training pruning of Stable Diffusion 2, addressing the critical need for model compression in text-to-image domain. Our study tackles the pruning techniques for the previously unexplored multi-modal generation models, and particularly examines the pruning impact on the textual component and the image generation component separately. We conduct a comprehensive comparison on pruning the model or the single component of the model in various sparsities. Our results yield previously undocumented findings. For example, contrary to established trends in language model pruning, we discover that simple magnitude pruning outperforms more advanced techniques in text-to-image context. Furthermore, our results show that Stable Diffusion 2 can be pruned to 38.5% sparsity with minimal quality loss, achieving a significant reduction in model size. We propose an optimal pruning configuration that prunes the text encoder to 47.5% and the diffusion generator to 35%. This configuration maintains image generation quality while substantially reducing computational requirements. In addition, our work uncovers intriguing questions about information encoding in text-to-image models: we observe that pruning beyond certain thresholds leads to sudden performance drops (unreadable images), suggesting that specific weights encode critical semantics information. This finding opens new avenues for future research in model compression, interoperability, and bias identification in text-to-image models. By providing crucial insights into the pruning behavior of text-to-image models, our study lays the groundwork for developing more efficient and accessible AI-driven image generation systems
Authors: Ri-Zhao Qiu, Yuchen Song, Xuanbin Peng, Sai Aneesh Suryadevara, Ge Yang, Minghuan Liu, Mazeyu Ji, Chengzhe Jia, Ruihan Yang, Xueyan Zou, Xiaolong Wang
Abstract: `In-the-wild' mobile manipulation aims to deploy robots in diverse real-world environments, which requires the robot to (1) have skills that generalize across object configurations; (2) be capable of long-horizon task execution in diverse environments; and (3) perform complex manipulation beyond pick-and-place. Quadruped robots with manipulators hold promise for extending the workspace and enabling robust locomotion, but existing results do not investigate such a capability. This paper proposes WildLMa with three components to address these issues: (1) adaptation of learned low-level controller for VR-enabled whole-body teleoperation and traversability; (2) WildLMa-Skill -- a library of generalizable visuomotor skills acquired via imitation learning or heuristics and (3) WildLMa-Planner -- an interface of learned skills that allow LLM planners to coordinate skills for long-horizon tasks. We demonstrate the importance of high-quality training data by achieving higher grasping success rate over existing RL baselines using only tens of demonstrations. WildLMa exploits CLIP for language-conditioned imitation learning that empirically generalizes to objects unseen in training demonstrations. Besides extensive quantitative evaluation, we qualitatively demonstrate practical robot applications, such as cleaning up trash in university hallways or outdoor terrains, operating articulated objects, and rearranging items on a bookshelf.
Authors: Niharika Shimona D'Souza, Mary Beth Nebel, Deana Crocetti, Nicholas Wymbs, Joshua Robinson, Stewart Mostofsky, Archana Venkataraman
Abstract: We propose an integrated deep-generative framework, that jointly models complementary information from resting-state functional MRI (rs-fMRI) connectivity and diffusion tensor imaging (DTI) tractography to extract predictive biomarkers of a disease. The generative part of our framework is a structurally-regularized Dynamic Dictionary Learning (sr-DDL) model that decomposes the dynamic rs-fMRI correlation matrices into a collection of shared basis networks and time varying patient-specific loadings. This matrix factorization is guided by the DTI tractography matrices to learn anatomically informed connectivity profiles. The deep part of our framework is an LSTM-ANN block, which models the temporal evolution of the patient sr-DDL loadings to predict multidimensional clinical severity. Our coupled optimization procedure collectively estimates the basis networks, the patient-specific dynamic loadings, and the neural network weights. We validate our framework on a multi-score prediction task in 57 patients diagnosed with Autism Spectrum Disorder (ASD). Our hybrid model outperforms state-of-the-art baselines in a five-fold cross validated setting and extracts interpretable multimodal neural signatures of brain dysfunction in ASD.
Authors: Vlad Niculae, Caio F. Corro, Nikita Nangia, Tsvetomila Mihaylova, Andr\'e F. T. Martins
Abstract: Many types of data from fields including natural language processing, computer vision, and bioinformatics, are well represented by discrete, compositional structures such as trees, sequences, or matchings. Latent structure models are a powerful tool for learning to extract such representations, offering a way to incorporate structural bias, discover insight about the data, and interpret decisions. However, effective training is challenging, as neural networks are typically designed for continuous computation. This text explores three broad strategies for learning with discrete latent structure: continuous relaxation, surrogate gradients, and probabilistic estimation. Our presentation relies on consistent notations for a wide range of models. As such, we reveal many new connections between latent structure learning strategies, showing how most consist of the same small set of fundamental building blocks, but use them differently, leading to substantially different applicability and properties.
Authors: Chenggang Zhao, Genghan Zhang, Ao Shen, Mingyu Gao
Abstract: The demands for higher performance and accuracy in neural networks (NNs) never end. Existing tensor compilation and Neural Architecture Search (NAS) techniques orthogonally optimize the two goals but actually share many similarities in their concrete strategies. We exploit such opportunities by combining the two into one and make a case for Kernel Architecture Search (KAS). KAS reviews NAS from a system perspective and zooms into a more fine-grained level to generate neural kernels with both high performance and good accuracy. To demonstrate the potential of KAS, we build an end-to-end framework, Canvas, to find high-quality kernels as convolution replacements. Canvas samples from a rich set of fine-grained primitives to stochastically and iteratively construct new kernels and evaluate them according to user-specified constraints. Canvas supports freely adjustable tensor dimension sizes inside the kernel and uses two levels of solvers to satisfy structural legality and fully utilize model budgets. The evaluation shows that by replacing standard convolutions with generated new kernels in common NNs, Canvas achieves average 1.5x speedups compared to the previous state-of-the-art with acceptable accuracy loss and search efficiency. Canvas verifies the practicability of KAS by rediscovering many manually designed kernels in the past and producing new structures that may inspire future machine learning innovations. For source code and implementation, we open-sourced Canvas at https://github.com/tsinghua-ideal/Canvas.
Authors: Giannis Nikolentzos, Michail Chatzianastasis, Michalis Vazirgiannis
Abstract: In recent years, graph neural networks (GNNs) have achieved great success in the field of graph representation learning. Although prior work has shed light on the expressiveness of those models (\ie whether they can distinguish pairs of non-isomorphic graphs), it is still not clear what structural information is encoded into the node representations that are learned by those models. In this paper, we address this gap by studying the node representations learned by four standard GNN models. We find that some models produce identical representations for all nodes, while the representations learned by other models are linked to some notion of walks of specific length that start from the nodes. We establish Lipschitz bounds for these models with respect to the number of (normalized) walks. Additionally, we investigate the influence of node features on the learned representations. We find that if the initial representations of all nodes point in the same direction, the representations learned at the $k$-th layer of the models are also related to the initial features of nodes that can be reached in exactly $k$ steps. We also apply our findings to understand the phenomenon of oversquashing that occurs in GNNs. Our theoretical analysis is validated through experiments on synthetic and real-world datasets.
Authors: Zhanwei Wang, Kaibin Huang, Yonina C. Eldar
Abstract: Federated Learning (FL) is a widely embraced paradigm for distilling artificial intelligence from distributed mobile data. However, the deployment of FL in mobile networks can be compromised by exposure to interference from neighboring cells or jammers. Existing interference mitigation techniques require multi-cell cooperation or at least interference channel state information, which is expensive in practice. On the other hand, power control that treats interference as noise may not be effective due to limited power budgets, and also that this mechanism can trigger countermeasures by interference sources. As a practical approach for protecting FL against interference, we propose Spectrum Breathing, which cascades stochastic-gradient pruning and spread spectrum to suppress interference without bandwidth expansion. The cost is higher learning latency by exploiting the graceful degradation of learning speed due to pruning. We synchronize the two operations such that their levels are controlled by the same parameter, Breathing Depth. To optimally control the parameter, we develop a martingale-based approach to convergence analysis of Over-the-Air FL with spectrum breathing, termed AirBreathing FL. We show a performance tradeoff between gradient-pruning and interference-induced error as regulated by the breathing depth. Given receive SIR and model size, the optimization of the tradeoff yields two schemes for controlling the breathing depth that can be either fixed or adaptive to channels and the learning process. As shown by experiments, in scenarios where traditional Over-the-Air FL fails to converge in the presence of strong interference, AirBreahing FL with either fixed or adaptive breathing depth can ensure convergence where the adaptive scheme achieves close-to-ideal performance.
Authors: Feng-Lei Fan, Wei Huang, Xiangru Zhong, Lecheng Ruan, Tieyong Zeng, Huan Xiong, Fei Wang
Abstract: A ReLU network is a piecewise linear function over polytopes. Figuring out the properties of such polytopes is of fundamental importance for the research and development of neural networks. So far, either theoretical or empirical studies on polytopes only stay at the level of counting their number, which is far from a complete characterization. Here, we propose to study the shapes of polytopes via the number of faces of the polytope. Then, by computing and analyzing the histogram of faces across polytopes, we find that a ReLU network has relatively simple polytopes under both initialization and gradient descent, although these polytopes can be rather diverse and complicated by a specific design. This finding can be appreciated as a kind of generalized implicit bias, subjected to the intrinsic geometric constraint in space partition of a ReLU network. Next, we perform a combinatorial analysis to explain why adding depth does not generate a more complicated polytope by bounding the average number of faces of polytopes with the dimensionality. Our results concretely reveal what kind of simple functions a network learns and what will happen when a network goes deep. Also, by characterizing the shape of polytopes, the number of faces can be a novel leverage for other problems, \textit{e.g.}, serving as a generic tool to explain the power of popular shortcut networks such as ResNet and analyzing the impact of different regularization strategies on a network's space partition.
Authors: Anwar Said, Roza G. Bayrak, Tyler Derr, Mudassir Shabbir, Daniel Moyer, Catie Chang, Xenofon Koutsoukos
Abstract: Machine learning provides a valuable tool for analyzing high-dimensional functional neuroimaging data, and is proving effective in predicting various neurological conditions, psychiatric disorders, and cognitive patterns. In functional magnetic resonance imaging (MRI) research, interactions between brain regions are commonly modeled using graph-based representations. The potency of graph machine learning methods has been established across myriad domains, marking a transformative step in data interpretation and predictive modeling. Yet, despite their promise, the transposition of these techniques to the neuroimaging domain has been challenging due to the expansive number of potential preprocessing pipelines and the large parameter search space for graph-based dataset construction. In this paper, we introduce NeuroGraph, a collection of graph-based neuroimaging datasets, and demonstrated its utility for predicting multiple categories of behavioral and cognitive traits. We delve deeply into the dataset generation search space by crafting 35 datasets that encompass static and dynamic brain connectivity, running in excess of 15 baseline methods for benchmarking. Additionally, we provide generic frameworks for learning on both static and dynamic graphs. Our extensive experiments lead to several key observations. Notably, using correlation vectors as node features, incorporating larger number of regions of interest, and employing sparser graphs lead to improved performance. To foster further advancements in graph-based data driven neuroimaging analysis, we offer a comprehensive open-source Python package that includes the benchmark datasets, baseline implementations, model training, and standard evaluation.
Authors: Santiago del Rey, Silverio Mart\'inez-Fern\'andez, Lu\'is Cruz, Xavier Franch
Abstract: To raise awareness of the huge impact Deep Learning (DL) has on the environment, several works have tried to estimate the energy consumption and carbon footprint of DL-based systems across their life cycle. However, the estimations for energy consumption in the training stage usually rely on assumptions that have not been thoroughly tested. This study aims to move past these assumptions by leveraging the relationship between energy consumption and two relevant design decisions in DL training; model architecture, and training environment. To investigate these relationships, we collect multiple metrics related to energy efficiency and model correctness during the models' training. Then, we outline the trade-offs between the measured energy consumption and the models' correctness regarding model architecture, and their relationship with the training environment. Finally, we study the training's power consumption behavior and propose four new energy estimation methods. Our results show that selecting the proper model architecture and training environment can reduce energy consumption dramatically (up to 80.72%) at the cost of negligible decreases in correctness. Also, we find evidence that GPUs should scale with the models' computational complexity for better energy efficiency. Furthermore, we prove that current energy estimation methods are unreliable and propose alternatives 2x more precise.
Authors: Xixi Hu, Bo Liu, Xingchao Liu, Qiang Liu
Abstract: Diffusion-based imitation learning improves Behavioral Cloning (BC) on multi-modal decision-making, but comes at the cost of significantly slower inference due to the recursion in the diffusion process. It urges us to design efficient policy generators while keeping the ability to generate diverse actions. To address this challenge, we propose AdaFlow, an imitation learning framework based on flow-based generative modeling. AdaFlow represents the policy with state-conditioned ordinary differential equations (ODEs), which are known as probability flows. We reveal an intriguing connection between the conditional variance of their training loss and the discretization error of the ODEs. With this insight, we propose a variance-adaptive ODE solver that can adjust its step size in the inference stage, making AdaFlow an adaptive decision-maker, offering rapid inference without sacrificing diversity. Interestingly, it automatically reduces to a one-step generator when the action distribution is uni-modal. Our comprehensive empirical evaluation shows that AdaFlow achieves high performance with fast inference speed.
Authors: YongKyung Oh, Dongyoung Lim, Sungil Kim
Abstract: Irregular sampling intervals and missing values in real-world time series data present challenges for conventional methods that assume consistent intervals and complete data. Neural Ordinary Differential Equations (Neural ODEs) offer an alternative approach, utilizing neural networks combined with ODE solvers to learn continuous latent representations through parameterized vector fields. Neural Stochastic Differential Equations (Neural SDEs) extend Neural ODEs by incorporating a diffusion term, although this addition is not trivial, particularly when addressing irregular intervals and missing values. Consequently, careful design of drift and diffusion functions is crucial for maintaining stability and enhancing performance, while incautious choices can result in adverse properties such as the absence of strong solutions, stochastic destabilization, or unstable Euler discretizations, significantly affecting Neural SDEs' performance. In this study, we propose three stable classes of Neural SDEs: Langevin-type SDE, Linear Noise SDE, and Geometric SDE. Then, we rigorously demonstrate their robustness in maintaining excellent performance under distribution shift, while effectively preventing overfitting. To assess the effectiveness of our approach, we conduct extensive experiments on four benchmark datasets for interpolation, forecasting, and classification tasks, and analyze the robustness of our methods with 30 public datasets under different missing rates. Our results demonstrate the efficacy of the proposed method in handling real-world irregular time series data.
Authors: Nizar Masmoudi, Wael Jaafar
Abstract: The conjunction of edge intelligence and the ever-growing Internet-of-Things (IoT) network heralds a new era of collaborative machine learning, with federated learning (FL) emerging as the most prominent paradigm. With the growing interest in these learning schemes, researchers started addressing some of their most fundamental limitations. Indeed, conventional FL with a central aggregator presents a single point of failure and a network bottleneck. To bypass this issue, decentralized FL where nodes collaborate in a peer-to-peer network has been proposed. Despite the latter's efficiency, communication costs and data heterogeneity remain key challenges in decentralized FL. In this context, we propose a novel scheme, called opportunistic communication-efficient decentralized federated learning, a.k.a., OCD-FL, consisting of a systematic FL peer selection for collaboration, aiming to achieve maximum FL knowledge gain while reducing energy consumption. Experimental results demonstrate the capability of OCD-FL to achieve similar or better performances than the fully collaborative FL, while significantly reducing consumed energy by at least 30% and up to 80%.
Authors: Zitao Shuai, Chenwei Wu, Zhengxu Tang, Liyue Shen
Abstract: Vision-language pre-training (VLP) has emerged as an effective scheme for multimodal representation learning, but its reliance on large-scale multimodal data poses significant challenges for medical applications. Federated learning (FL) offers a promising solution to scale up the dataset for medical VLP while preserving data privacy. However, we observe that client data heterogeneity in real-world scenarios could cause models to learn biased cross-modal alignment during local pre-training. This would limit the transferability of the federally learned representation model on downstream tasks. To address this challenge, we propose Federated Distributionally Robust Alignment (FedDRA), a framework for federated VLP that achieves robust vision-language alignment under heterogeneous conditions. Based on client datasets, we construct a distribution family that encompasses potential test-time domains, and apply a distributionally robust framework to optimize the pre-trained model's performance across this distribution space. This approach bridges the gap between pre-training samples and downstream applications. To avoid over-fitting on client-specific information, we use anchor representation from the global model to guide the local training, and adopt a two-stage approach to first tune deeper layers before updating the entire network. Extensive experiments on real-world datasets demonstrate FedDRA's effectiveness in enhancing medical federated VLP under data heterogeneity. Our method also adapts well to various medical pre-training methods.
Authors: Junqi Jiang, Francesco Leofante, Antonio Rago, Francesca Toni
Abstract: Counterfactual Explanations (CEs) have emerged as a major paradigm in explainable AI research, providing recourse recommendations for users affected by the decisions of machine learning models. However, CEs found by existing methods often become invalid when slight changes occur in the parameters of the model they were generated for. The literature lacks a way to provide exhaustive robustness guarantees for CEs under model changes, in that existing methods to improve CEs' robustness are mostly heuristic, and the robustness performances are evaluated empirically using only a limited number of retrained models. To bridge this gap, we propose a novel interval abstraction technique for parametric machine learning models, which allows us to obtain provable robustness guarantees for CEs under a possibly infinite set of plausible model changes $\Delta$. Based on this idea, we formalise a robustness notion for CEs, which we call $\Delta$-robustness, in both binary and multi-class classification settings. We present procedures to verify $\Delta$-robustness based on Mixed Integer Linear Programming, using which we further propose algorithms to generate CEs that are $\Delta$-robust. In an extensive empirical study involving neural networks and logistic regression models, we demonstrate the practical applicability of our approach. We discuss two strategies for determining the appropriate hyperparameters in our method, and we quantitatively benchmark CEs generated by eleven methods, highlighting the effectiveness of our algorithms in finding robust CEs.
Authors: Chenghao Huang, Xiaolu Chen, Yanru Zhang, Hao Wang
Abstract: Heterogeneity resulting from label distribution skew and data scarcity can lead to inaccuracy and unfairness in intelligent communication applications that mainly rely on distributed computing. To deal with it, this paper proposes a novel personalized federated learning algorithm, named Federated Contrastive Shareable Representations (FedCoSR), to facilitate knowledge sharing among clients while maintaining data privacy. Specifically, parameters of local models' shallow layers and typical local representations are both considered shareable information for the server and aggregated globally. To address poor performance caused by label distribution skew among clients, contrastive learning is adopted between local and global representations to enrich local knowledge. Additionally, to ensure fairness for clients with scarce data, FedCoSR introduces adaptive local aggregation to coordinate the global model involvement in each client. Our simulations demonstrate FedCoSR's effectiveness in mitigating label heterogeneity by achieving accuracy and fairness improvements over existing methods on datasets with varying degrees of label heterogeneity.
Authors: Michal Nauman, Mateusz Ostaszewski, Krzysztof Jankowski, Piotr Mi{\l}o\'s, Marek Cygan
Abstract: Sample efficiency in Reinforcement Learning (RL) has traditionally been driven by algorithmic enhancements. In this work, we demonstrate that scaling can also lead to substantial improvements. We conduct a thorough investigation into the interplay of scaling model capacity and domain-specific RL enhancements. These empirical findings inform the design choices underlying our proposed BRO (Bigger, Regularized, Optimistic) algorithm. The key innovation behind BRO is that strong regularization allows for effective scaling of the critic networks, which, paired with optimistic exploration, leads to superior performance. BRO achieves state-of-the-art results, significantly outperforming the leading model-based and model-free algorithms across 40 complex tasks from the DeepMind Control, MetaWorld, and MyoSuite benchmarks. BRO is the first model-free algorithm to achieve near-optimal policies in the notoriously challenging Dog and Humanoid tasks.
Authors: Qingyun Sun, Ziying Chen, Beining Yang, Cheng Ji, Xingcheng Fu, Sheng Zhou, Hao Peng, Jianxin Li, Philip S. Yu
Abstract: Graph condensation (GC) has recently garnered considerable attention due to its ability to reduce large-scale graph datasets while preserving their essential properties. The core concept of GC is to create a smaller, more manageable graph that retains the characteristics of the original graph. Despite the proliferation of graph condensation methods developed in recent years, there is no comprehensive evaluation and in-depth analysis, which creates a great obstacle to understanding the progress in this field. To fill this gap, we develop a comprehensive Graph Condensation Benchmark (GC-Bench) to analyze the performance of graph condensation in different scenarios systematically. Specifically, GC-Bench systematically investigates the characteristics of graph condensation in terms of the following dimensions: effectiveness, transferability, and complexity. We comprehensively evaluate 12 state-of-the-art graph condensation algorithms in node-level and graph-level tasks and analyze their performance in 12 diverse graph datasets. Further, we have developed an easy-to-use library for training and evaluating different GC methods to facilitate reproducible research. The GC-Bench library is available at https://github.com/RingBDStack/GC-Bench.
Authors: Baiyu Peng, Aude Billard
Abstract: Planning for a wide range of real-world tasks necessitates to know and write all constraints. However, instances exist where these constraints are either unknown or challenging to specify accurately. A possible solution is to infer the unknown constraints from expert demonstration. The majority of prior works limit themselves to learning simple linear constraints, or require strong knowledge of the true constraint parameterization or environmental model. To mitigate these problems, this paper presents a positive-unlabeled (PU) learning approach to infer a continuous, arbitrary and possibly nonlinear, constraint from demonstration. From a PU learning view, We treat all data in demonstrations as positive (feasible) data, and learn a (sub)-optimal policy to generate high-reward-winning but potentially infeasible trajectories, which serve as unlabeled data containing both feasible and infeasible states. Under an assumption on data distribution, a feasible-infeasible classifier (i.e., constraint model) is learned from the two datasets through a postprocessing PU learning technique. The entire method employs an iterative framework alternating between updating the policy, which generates and selects higher-reward policies, and updating the constraint model. Additionally, a memory buffer is introduced to record and reuse samples from previous iterations to prevent forgetting. The effectiveness of the proposed method is validated in two Mujoco environments, successfully inferring continuous nonlinear constraints and outperforming a baseline method in terms of constraint accuracy and policy safety.
Authors: Wall Kim
Abstract: Sequence modeling with State Space models (SSMs) has demonstrated performance surpassing that of Transformers in various tasks, raising expectations for their potential to outperform the Decision Transformer and its enhanced variants in offline reinforcement learning (RL). However, decision models based on Mamba, a state-of-the-art SSM, failed to achieve superior performance compared to these enhanced Decision Transformers. We hypothesize that this limitation arises from information loss during the selective scanning phase. To address this, we propose the Decision MetaMamba (DMM), which augments Mamba with a token mixer in its input layer. This mixer explicitly accounts for the multimodal nature of offline RL inputs, comprising state, action, and return-to-go. The DMM demonstrates improved performance while significantly reducing parameter count compared to prior models. Notably, similar performance gains were achieved using a simple linear token mixer, emphasizing the importance of preserving information from proximate time steps rather than the specific design of the token mixer itself. This novel modification to Mamba's input layer represents a departure from conventional timestamp-based encoding approaches used in Transformers. By enhancing performance of Mamba in offline RL, characterized by memory efficiency and fast inference, this work opens new avenues for its broader application in future RL research.
Authors: Yan Wu, Esther Wershof, Sebastian M Schmon, Marcel Nassar, B{\l}a\.zej Osi\'nski, Ridvan Eksi, Kun Zhang, Thore Graepel
Abstract: We present a comprehensive framework for predicting the effects of perturbations in single cells, designed to standardize benchmarking in this rapidly evolving field. Our framework, PerturBench, includes a user-friendly platform, diverse datasets, metrics for fair model comparison, and detailed performance analysis. Extensive evaluations of published and baseline models reveal limitations like mode or posterior collapse, and underscore the importance of rank metrics that assess the ordering of perturbations alongside traditional measures like RMSE. Our findings show that simple models can outperform more complex approaches. This benchmarking exercise sets new standards for model evaluation, supports robust model development, and advances the potential of these models to use high-throughput and high-content genetic and chemical screens for disease target discovery.
Authors: Shervin Khalafi, Dongsheng Ding, Alejandro Ribeiro
Abstract: Diffusion models have attained prominence for their ability to synthesize a probability distribution for a given dataset via a diffusion process, enabling the generation of new data points with high fidelity. However, diffusion processes are prone to generating samples that reflect biases in a training dataset. To address this issue, we develop constrained diffusion models by imposing diffusion constraints based on desired distributions that are informed by requirements. Specifically, we cast the training of diffusion models under requirements as a constrained distribution optimization problem that aims to reduce the distribution difference between original and generated data while obeying constraints on the distribution of generated data. We show that our constrained diffusion models generate new data from a mixture data distribution that achieves the optimal trade-off among objective and constraints. To train constrained diffusion models, we develop a dual training algorithm and characterize the optimality of the trained constrained diffusion model. We empirically demonstrate the effectiveness of our constrained models in two constrained generation tasks: (i) we consider a dataset with one or more underrepresented classes where we train the model with constraints to ensure fairly sampling from all classes during inference; (ii) we fine-tune a pre-trained diffusion model to sample from a new dataset while avoiding overfitting.
Authors: Jiarui Jiang, Wei Huang, Miao Zhang, Taiji Suzuki, Liqiang Nie
Abstract: Transformers have demonstrated great power in the recent development of large foundational models. In particular, the Vision Transformer (ViT) has brought revolutionary changes to the field of vision, achieving significant accomplishments on the experimental side. However, their theoretical capabilities, particularly in terms of generalization when trained to overfit training data, are still not fully understood. To address this gap, this work delves deeply into the benign overfitting perspective of transformers in vision. To this end, we study the optimization of a Transformer composed of a self-attention layer with softmax followed by a fully connected layer under gradient descent on a certain data distribution model. By developing techniques that address the challenges posed by softmax and the interdependent nature of multiple weights in transformer optimization, we successfully characterized the training dynamics and achieved generalization in post-training. Our results establish a sharp condition that can distinguish between the small test error phase and the large test error regime, based on the signal-to-noise ratio in the data model. The theoretical results are further verified by experimental simulation. To the best of our knowledge, this is the first work to characterize benign overfitting for Transformers.
Authors: Dylan Xu, Juan-Pablo Rivera
Abstract: Recent advances in deep learning have brought attention to the possibility of creating advanced, general AI systems that outperform humans across many tasks. However, if these systems pursue unintended goals, there could be catastrophic consequences. A key prerequisite for AI systems pursuing unintended goals is whether they will behave in a coherent and goal-directed manner in the first place, optimizing for some unknown goal; there exists significant research trying to evaluate systems for said behaviors. However, the most rigorous definitions of goal-directedness we currently have are difficult to compute in real-world settings. Drawing upon this previous literature, we explore policy goal-directedness within reinforcement learning (RL) environments. In our findings, we propose a different family of definitions of the goal-directedness of a policy that analyze whether it is well-modeled as near-optimal for many (sparse) reward functions. We operationalize this preliminary definition of goal-directedness and test it in toy Markov decision process (MDP) environments. Furthermore, we explore how goal-directedness could be measured in frontier large-language models (LLMs). Our contribution is a definition of goal-directedness that is simpler and more easily computable in order to approach the question of whether AI systems could pursue dangerous goals. We recommend further exploration of measuring coherence and goal-directedness, based on our findings.
Authors: Zirui Li, Patrick Wolf, Meng Wang
Abstract: Accurate arrival time prediction (ATP) of buses and trams plays a crucial role in public transport operations. Current methods focused on modeling one-dimensional temporal information but overlooked the latent periodic information within time series. Moreover, most studies developed algorithms for ATP based on a single or a few routes of public transport, which reduces the transferability of the prediction models and their applicability in public transport management systems. To this end, this paper proposes \textit{ArrivalNet}, a two-dimensional temporal variation-based multi-step ATP for buses and trams. It decomposes the one-dimensional temporal sequence into intra-periodic and inter-periodic variations, which can be recast into two-dimensional tensors (2D blocks). Each row of a tensor contains the time points within a period, and each column involves the time points at the same intra-periodic index across various periods. The transformed 2D blocks in different frequencies have an image-like feature representation that enables effective learning with computer vision backbones (e.g., convolutional neural network). Drawing on the concept of residual neural network, the 2D block module is designed as a basic module for flexible aggregation. Meanwhile, contextual factors like workdays, peak hours, and intersections, are also utilized in the augmented feature representation to improve the performance of prediction. 125 days of public transport data from Dresden were collected for model training and validation. Experimental results show that the root mean square error, mean absolute error, and mean absolute percentage error of the proposed predictor decrease by at least 6.1\%, 14.7\%, and 34.2\% compared with state-of-the-art baseline methods.
Authors: Ahmed Ala Eddine Benali, Massimo Cafaro, Italo Epicoco, Marco Pulimeno, Enrico Junior Schioppa
Abstract: Precise energy load forecasting in residential households is crucial for mitigating carbon emissions and enhancing energy efficiency; indeed, accurate forecasting enables utility companies and policymakers, who advocate sustainable energy practices, to optimize resource utilization. Moreover, smart meters provide valuable information by allowing for granular insights into consumption patterns. Building upon available smart meter data, our study aims to cluster consumers into distinct groups according to their energy usage behaviours, effectively capturing a diverse spectrum of consumption patterns. Next, we design JITtrans (Just In Time transformer), a novel transformer deep learning model that significantly improves energy consumption forecasting accuracy, with respect to traditional forecasting methods. Extensive experimental results validate our claims using proprietary smart meter data. Our findings highlight the potential of advanced predictive technologies to revolutionize energy management and advance sustainable power systems: the development of efficient and eco-friendly energy solutions critically depends on such technologies.
Authors: Bong Gyun Kang, Dongjun Lee, HyunGi Kim, DoHyun Chung, Sungroh Yoon
Abstract: Sequence modeling faces challenges in capturing long-range dependencies across diverse tasks. Recent linear and transformer-based forecasters have shown superior performance in time series forecasting. However, they are constrained by their inherent inability to effectively address long-range dependencies in time series data, primarily due to using fixed-size inputs for prediction. Furthermore, they typically sacrifice essential temporal correlation among consecutive training samples by shuffling them into mini-batches. To overcome these limitations, we introduce a fast and effective Spectral Attention mechanism, which preserves temporal correlations among samples and facilitates the handling of long-range information while maintaining the base model structure. Spectral Attention preserves long-period trends through a low-pass filter and facilitates gradient to flow between samples. Spectral Attention can be seamlessly integrated into most sequence models, allowing models with fixed-sized look-back windows to capture long-range dependencies over thousands of steps. Through extensive experiments on 11 real-world time series datasets using 7 recent forecasting models, we consistently demonstrate the efficacy of our Spectral Attention mechanism, achieving state-of-the-art results.
Authors: Pau Rodriguez, Arno Blaas, Michal Klein, Luca Zappella, Nicholas Apostoloff, Marco Cuturi, Xavier Suau
Abstract: The increasing capabilities of large generative models and their ever more widespread deployment have raised concerns about their reliability, safety, and potential misuse. To address these issues, recent works have proposed to control model generation by steering model activations in order to effectively induce or prevent the emergence of concepts or behaviors in the generated output. In this paper we introduce Activation Transport (AcT), a general framework to steer activations guided by optimal transport theory that generalizes many previous activation-steering works. AcT is modality-agnostic and provides fine-grained control over the model behavior with negligible computational overhead, while minimally impacting model abilities. We experimentally show the effectiveness and versatility of our approach by addressing key challenges in large language models (LLMs) and text-to-image diffusion models (T2Is). For LLMs, we show that AcT can effectively mitigate toxicity, induce arbitrary concepts, and increase their truthfulness. In T2Is, we show how AcT enables fine-grained style control and concept negation.
Authors: Wenxuan Li, Taiyi Wang, Eiko Yoneki
Abstract: Optimizing black-box functions in high-dimensional search spaces has been known to be challenging for traditional Bayesian Optimization (BO). In this paper, we introduce HiBO, a novel hierarchical algorithm integrating global-level search space partitioning information into the acquisition strategy of a local BO-based optimizer. HiBO employs a search-tree-based global-level navigator to adaptively split the search space into partitions with different sampling potential. The local optimizer then utilizes this global-level information to guide its acquisition strategy towards most promising regions within the search space. A comprehensive set of evaluations demonstrates that HiBO outperforms state-of-the-art methods in high-dimensional synthetic benchmarks and presents significant practical effectiveness in the real-world task of tuning configurations of database management systems (DBMSs).
Authors: Xiang Li, Yixiang Dai, Qing Qu
Abstract: In this work, we study the generalizability of diffusion models by looking into the hidden properties of the learned score functions, which are essentially a series of deep denoisers trained on various noise levels. We observe that as diffusion models transition from memorization to generalization, their corresponding nonlinear diffusion denoisers exhibit increasing linearity. This discovery leads us to investigate the linear counterparts of the nonlinear diffusion models, which are a series of linear models trained to match the function mappings of the nonlinear diffusion denoisers. Surprisingly, these linear denoisers are approximately the optimal denoisers for a multivariate Gaussian distribution characterized by the empirical mean and covariance of the training dataset. This finding implies that diffusion models have the inductive bias towards capturing and utilizing the Gaussian structure (covariance information) of the training dataset for data generation. We empirically demonstrate that this inductive bias is a unique property of diffusion models in the generalization regime, which becomes increasingly evident when the model's capacity is relatively small compared to the training dataset size. In the case that the model is highly overparameterized, this inductive bias emerges during the initial training phases before the model fully memorizes its training data. Our study provides crucial insights into understanding the notable strong generalization phenomenon recently observed in real-world diffusion models.
Authors: Shohei Taniguchi, Keno Harada, Gouki Minegishi, Yuta Oshima, Seong Cheol Jeong, Go Nagahara, Tomoshi Iiyama, Masahiro Suzuki, Yusuke Iwasawa, Yutaka Matsuo
Abstract: Adam is one of the most popular optimization algorithms in deep learning. However, it is known that Adam does not converge in theory unless choosing a hyperparameter, i.e., $\beta_2$, in a problem-dependent manner. There have been many attempts to fix the non-convergence (e.g., AMSGrad), but they require an impractical assumption that the gradient noise is uniformly bounded. In this paper, we propose a new adaptive gradient method named ADOPT, which achieves the optimal convergence rate of $\mathcal{O} ( 1 / \sqrt{T} )$ with any choice of $\beta_2$ without depending on the bounded noise assumption. ADOPT addresses the non-convergence issue of Adam by removing the current gradient from the second moment estimate and changing the order of the momentum update and the normalization by the second moment estimate. We also conduct intensive numerical experiments, and verify that our ADOPT achieves superior results compared to Adam and its variants across a wide range of tasks, including image classification, generative modeling, natural language processing, and deep reinforcement learning. The implementation is available at https://github.com/iShohei220/adopt.
Authors: Prashank Kadam
Abstract: As the availability of financial services online continues to grow, the incidence of fraud has surged correspondingly. Fraudsters continually seek new and innovative ways to circumvent the detection algorithms in place. Traditionally, fraud detection relied on rule-based methods, where rules were manually created based on transaction data features. However, these techniques soon became ineffective due to their reliance on manual rule creation and their inability to detect complex data patterns. Today, a significant portion of the financial services sector employs various machine learning algorithms, such as XGBoost, Random Forest, and neural networks, to model transaction data. While these techniques have proven more efficient than rule-based methods, they still fail to capture interactions between different transactions and their interrelationships. Recently, graph-based techniques have been adopted for financial fraud detection, leveraging graph topology to aggregate neighborhood information of transaction data using Graph Neural Networks (GNNs). Despite showing improvements over previous methods, these techniques still struggle to keep pace with the evolving camouflaging tactics of fraudsters and suffer from information loss due to over-smoothing. In this paper, we propose a novel algorithm that employs an efficient neighborhood sampling method, effective for camouflage detection and preserving crucial feature information from non-similar nodes. Additionally, we introduce a novel GNN architecture that utilizes attention mechanisms and preserves holistic neighborhood information to prevent information loss. We test our algorithm on financial data to show that our method outperforms other state-of-the-art graph algorithms.
Authors: Jaehyeok Lee, Keisuke Sakaguchi, JinYeong Bak
Abstract: Self-training approach for large language models (LLMs) improves reasoning abilities by training the models on their self-generated rationales. Previous approaches have labeled rationales that produce correct answers for a given question as appropriate for training. However, a single measure risks misjudging rationale quality, leading the models to learn flawed reasoning patterns. To address this issue, we propose CREST (Consistency-driven Rationale Evaluation for Self-Training), a self-training framework that further evaluates each rationale through follow-up questions and leverages this evaluation to guide its training. Specifically, we introduce two methods: (1) filtering out rationales that frequently result in incorrect answers on follow-up questions and (2) preference learning based on mixed preferences from rationale evaluation results of both original and follow-up questions. Experiments on three question-answering datasets using open LLMs show that CREST not only improves the logical robustness and correctness of rationales but also improves reasoning abilities compared to previous self-training approaches.
Authors: Jinming Xing, Ruilin Xing
Abstract: Link prediction is crucial for understanding complex networks but traditional Graph Neural Networks (GNNs) often rely on random negative sampling, leading to suboptimal performance. This paper introduces Fuzzy Graph Attention Networks (FGAT), a novel approach integrating fuzzy rough sets for dynamic negative sampling and enhanced node feature aggregation. Fuzzy Negative Sampling (FNS) systematically selects high-quality negative edges based on fuzzy similarities, improving training efficiency. FGAT layer incorporates fuzzy rough set principles, enabling robust and discriminative node representations. Experiments on two research collaboration networks demonstrate FGAT's superior link prediction accuracy, outperforming state-of-the-art baselines by leveraging the power of fuzzy rough sets for effective negative sampling and node feature learning.
Authors: Allen Hao Huang
Abstract: Activation functions play a crucial role in introducing non-linearities to deep neural networks. We propose a novel approach to designing activation functions by focusing on their gradients and deriving the corresponding functions through integration. Our work introduces the Expanded Integral of the Exponential Linear Unit (xIELU), a trainable piecewise activation function derived by integrating trainable affine transformations applied on the ELU activation function. xIELU combines two key gradient properties: a trainable and linearly increasing gradient for positive inputs, similar to ReLU$^2$, and a trainable negative gradient flow for negative inputs, akin to xSiLU. Conceptually, xIELU can be viewed as extending ReLU$^2$ to effectively handle negative inputs. In experiments with 1.1B parameter Llama models trained on 126B tokens of FineWeb Edu, xIELU achieves lower perplexity compared to both ReLU$^2$ and SwiGLU when matched for the same compute cost and parameter count.
Authors: Erica Coppolillo, Federico Cinus, Marco Minici, Francesco Bonchi, Giuseppe Manco
Abstract: Large Language Models (LLMs) exhibit significant persuasion capabilities in one-on-one interactions, but their influence within social networks remains underexplored. This study investigates the potential social impact of LLMs in these environments, where interconnected users and complex opinion dynamics pose unique challenges. In particular, we address the following research question: can LLMs learn to generate meaningful content that maximizes user engagement on social networks? To answer this question, we define a pipeline to guide the LLM-based content generation which employs reinforcement learning with simulated feedback. In our framework, the reward is based on an engagement model borrowed from the literature on opinion dynamics and information propagation. Moreover, we force the text generated by the LLM to be aligned with a given topic and to satisfy a minimum fluency requirement. Using our framework, we analyze the capabilities and limitations of LLMs in tackling the given task, specifically considering the relative positions of the LLM as an agent within the social network and the distribution of opinions in the network on the given topic. Our findings show the full potential of LLMs in creating social engagement. Notable properties of our approach are that the learning procedure is adaptive to the opinion distribution of the underlying network and agnostic to the specifics of the engagement model, which is embedded as a plug-and-play component. In this regard, our approach can be easily refined for more complex engagement tasks and interventions in computational social science. The code used for the experiments is publicly available at https://anonymous.4open.science/r/EDCG/.
Authors: Anxin Guo, Aravindan Vijayaraghavan
Abstract: We consider the problem of learning an arbitrarily-biased ReLU activation (or neuron) over Gaussian marginals with the squared loss objective. Despite the ReLU neuron being the basic building block of modern neural networks, we still do not understand the basic algorithmic question of whether one arbitrary ReLU neuron is learnable in the non-realizable setting. In particular, all existing polynomial time algorithms only provide approximation guarantees for the better-behaved unbiased setting or restricted bias setting. Our main result is a polynomial time statistical query (SQ) algorithm that gives the first constant factor approximation for arbitrary bias. It outputs a ReLU activation that achieves a loss of $O(\mathrm{OPT}) + \varepsilon$ in time $\mathrm{poly}(d,1/\varepsilon)$, where $\mathrm{OPT}$ is the loss obtained by the optimal ReLU activation. Our algorithm presents an interesting departure from existing algorithms, which are all based on gradient descent and thus fall within the class of correlational statistical query (CSQ) algorithms. We complement our algorithmic result by showing that no polynomial time CSQ algorithm can achieve a constant factor approximation. Together, these results shed light on the intrinsic limitation of gradient descent, while identifying arguably the simplest setting (a single neuron) where there is a separation between SQ and CSQ algorithms.
Authors: Wutao Wei, Nikhil Gupta, Bowei Xi
Abstract: Nowadays more and more data are gathered for detecting and preventing cyber attacks. In cyber security applications, data analytics techniques have to deal with active adversaries that try to deceive the data analytics models and avoid being detected. The existence of such adversarial behavior motivates the development of robust and resilient adversarial learning techniques for various tasks. Most of the previous work focused on adversarial classification techniques, which assumed the existence of a reasonably large amount of carefully labeled data instances. However, in practice, labeling the data instances often requires costly and time-consuming human expertise and becomes a significant bottleneck. Meanwhile, a large number of unlabeled instances can also be used to understand the adversaries' behavior. To address the above mentioned challenges, in this paper, we develop a novel grid based adversarial clustering algorithm. Our adversarial clustering algorithm is able to identify the core normal regions, and to draw defensive walls around the centers of the normal objects utilizing game theoretic ideas. Our algorithm also identifies sub-clusters of attack objects, the overlapping areas within clusters, and outliers which may be potential anomalies.
Authors: Niharika Shimona D'Souza, Mary Beth Nebel, Nicholas Wymbs, Stewart H. Mostofsky, Archana Venkataraman
Abstract: We propose a novel optimization framework to predict clinical severity from resting state fMRI (rs-fMRI) data. Our model consists of two coupled terms. The first term decomposes the correlation matrices into a sparse set of representative subnetworks that define a network manifold. These subnetworks are modeled as rank-one outer-products which correspond to the elemental patterns of co-activation across the brain; the subnetworks are combined via patient-specific non-negative coefficients. The second term is a linear regression model that uses the patient-specific coefficients to predict a measure of clinical severity. We validate our framework on two separate datasets in a ten fold cross validation setting. The first is a cohort of fifty-eight patients diagnosed with Autism Spectrum Disorder (ASD). The second dataset consists of sixty three patients from a publicly available ASD database. Our method outperforms standard semi-supervised frameworks, which employ conventional graph theoretic and statistical representation learning techniques to relate the rs-fMRI correlations to behavior. In contrast, our joint network optimization framework exploits the structure of the rs-fMRI correlation matrices to simultaneously capture group level effects and patient heterogeneity. Finally, we demonstrate that our proposed framework robustly identifies clinically relevant networks characteristic of ASD.
Authors: Giovanni S. Alberti, Matteo Santacesaria, Silvia Sciutto
Abstract: In this work, we present and study Continuous Generative Neural Networks (CGNNs), namely, generative models in the continuous setting: the output of a CGNN belongs to an infinite-dimensional function space. The architecture is inspired by DCGAN, with one fully connected layer, several convolutional layers and nonlinear activation functions. In the continuous $L^2$ setting, the dimensions of the spaces of each layer are replaced by the scales of a multiresolution analysis of a compactly supported wavelet. We present conditions on the convolutional filters and on the nonlinearity that guarantee that a CGNN is injective. This theory finds applications to inverse problems, and allows for deriving Lipschitz stability estimates for (possibly nonlinear) infinite-dimensional inverse problems with unknowns belonging to the manifold generated by a CGNN. Several numerical simulations, including signal deblurring, illustrate and validate this approach.
Authors: Gabriel Turinici
Abstract: We describe a measure quantization procedure i.e., an algorithm which finds the best approximation of a target probability law (and more generally signed finite variation measure) by a sum of $Q$ Dirac masses ($Q$ being the quantization parameter). The procedure is implemented by minimizing the statistical distance between the original measure and its quantized version; the distance is built from a negative definite kernel and, if necessary, can be computed on the fly and feed to a stochastic optimization algorithm (such as SGD, Adam, ...). We investigate theoretically the fundamental questions of existence of the optimal measure quantizer and identify what are the required kernel properties that guarantee suitable behavior. We propose two best linear unbiased (BLUE) estimators for the squared statistical distance and use them in an unbiased procedure, called HEMQ, to find the optimal quantization. We test HEMQ on several databases: multi-dimensional Gaussian mixtures, Wiener space cubature, Italian wine cultivars and the MNIST image database. The results indicate that the HEMQ algorithm is robust and versatile and, for the class of Huber-energy kernels, matches the expected intuitive behavior.
Authors: Adrian Arnaiz-Rodriguez, Georgina Curto, Nuria Oliver
Abstract: Social networks contribute to the distribution of social capital, defined as the relationships, norms of trust and reciprocity within a community or society that facilitate cooperation and collective action. Therefore, better positioned members in a social network benefit from faster access to diverse information and higher influence on information dissemination. A variety of methods have been proposed in the literature to measure social capital at an individual level. However, there is a lack of methods to quantify social capital at a group level, which is particularly important when the groups are defined on the grounds of protected attributes. To fill this gap, we propose to measure the social capital of a group of nodes by means of the effective resistance and emphasize the importance of considering the entire network topology. Grounded in spectral graph theory, we introduce three effective resistance-based measures of group social capital, namely group isolation, group diameter and group control, where the groups are defined according to the value of a protected attribute. We denote the social capital disparity among different groups in a network as structural group unfairness, and propose to mitigate it by means of a budgeted edge augmentation heuristic that systematically increases the social capital of the most disadvantaged group. In experiments on real-world networks, we uncover significant levels of structural group unfairness when using gender as the protected attribute, with females being the most disadvantaged group in comparison to males. We also illustrate how our proposed edge augmentation approach is able to not only effectively mitigate the structural group unfairness but also increase the social capital of all groups in the network.
Authors: Qiaozhe Zhang, Ruijie Zhang, Jun Sun, Yingzhuang Liu
Abstract: Network pruning is a commonly used measure to alleviate the storage and computational burden of deep neural networks. However, the fundamental limit of network pruning is still lacking. To close the gap, in this work we'll take a first-principles approach, i.e. we'll directly impose the sparsity constraint on the loss function and leverage the framework of statistical dimension in convex geometry, thus we're able to characterize the sharp phase transition point, i.e. the fundamental limit of the pruning ratio. Through this limit, we're able to identify two key factors that determine the pruning ratio limit, namely, weight magnitude and network sharpness. Generally speaking, the flatter the loss landscape or the smaller the weight magnitude, the smaller pruning ratio. Moreover, we provide efficient countermeasures to address the challenges in the computation of the pruning limit, which involves accurate spectrum estimation of a large-scale and non-positive Hessian matrix. Moreover, through the lens of the pruning ratio threshold, we can provide rigorous interpretations on several heuristics in existing pruning algorithms. Extensive experiments are performed that demonstrate that our theoretical pruning ratio threshold coincides very well with the experiments. All codes are available at: https://github.com/QiaozheZhang/Global-One-shot-Pruning
URLs: https://github.com/QiaozheZhang/Global-One-shot-Pruning
Authors: Haoran Liao, Derek S. Wang, Iskandar Sitdikov, Ciro Salcedo, Alireza Seif, Zlatko K. Minev
Abstract: Quantum computers progress toward outperforming classical supercomputers, but quantum errors remain their primary obstacle. The key to overcoming errors on near-term devices has emerged through the field of quantum error mitigation, enabling improved accuracy at the cost of additional run time. Here, through experiments on state-of-the-art quantum computers using up to 100 qubits, we demonstrate that without sacrificing accuracy machine learning for quantum error mitigation (ML-QEM) drastically reduces the cost of mitigation. We benchmark ML-QEM using a variety of machine learning models -- linear regression, random forests, multi-layer perceptrons, and graph neural networks -- on diverse classes of quantum circuits, over increasingly complex device-noise profiles, under interpolation and extrapolation, and in both numerics and experiments. These tests employ the popular digital zero-noise extrapolation method as an added reference. Finally, we propose a path toward scalable mitigation by using ML-QEM to mimic traditional mitigation methods with superior runtime efficiency. Our results show that classical machine learning can extend the reach and practicality of quantum error mitigation by reducing its overheads and highlight its broader potential for practical quantum computations.
Authors: Csaba Toth, Harald Oberhauser, Zoltan Szabo
Abstract: Tensor algebras give rise to one of the most powerful measures of similarity for sequences of arbitrary length called the signature kernel accompanied with attractive theoretical guarantees from stochastic analysis. Previous algorithms to compute the signature kernel scale quadratically in terms of the length and the number of the sequences. To mitigate this severe computational bottleneck, we develop a random Fourier feature-based acceleration of the signature kernel acting on the inherently non-Euclidean domain of sequences. We show uniform approximation guarantees for the proposed unbiased estimator of the signature kernel, while keeping its computation linear in the sequence length and number. In addition, combined with recent advances on tensor projections, we derive two even more scalable time series features with favourable concentration properties and computational complexity both in time and memory. Our empirical results show that the reduction in computational cost comes at a negligible price in terms of accuracy on moderate-sized datasets, and it enables one to scale to large datasets up to a million time series.
Authors: Linrui Dai, Rongzhao Zhang, Yongrui Yu, Xiaofan Zhang
Abstract: The recently emerging conditional diffusion models seem promising for mitigating the labor and expenses in building large 3D medical imaging datasets. However, previous studies on 3D CT generation have yet to fully capitalize on semantic and textual conditions, and they have primarily focused on specific organs characterized by a local structure and fixed contrast. In this work, we present GuideGen, a controllable framework that generates anatomical masks and corresponding CT volumes for the entire torso-from chest to pelvis-based on free-form text prompts. Our approach includes three core components: a text-conditional semantic synthesizer for creating realistic full-torso anatomies; a contrast-aware autoencoder for detailed, high-fidelity feature extraction across varying contrast levels; and a latent feature generator that ensures alignment between CT images, anatomical semantics and input prompts. To train and evaluate GuideGen, we compile a multi-modality cancer imaging dataset with paired CT and clinical descriptions from 12 public TCIA datasets and one private real-world dataset. Comprehensive evaluations across generation quality, cross-modality alignment, and data usability on multi-organ and tumor segmentation tasks demonstrate GuideGen's superiority over existing CT generation methods.
Authors: Gavin S. Hartnett, Aaron Barbosa, Pranav S. Mundada, Michael Hush, Michael J. Biercuk, Yuval Baum
Abstract: We introduce and experimentally test a machine-learning-based method for ranking logically equivalent quantum circuits based on expected performance estimates derived from a training procedure conducted on real hardware. We apply our method to the problem of layout selection, in which abstracted qubits are assigned to physical qubits on a given device. Circuit measurements performed on IBM hardware indicate that the maximum and median fidelities of logically equivalent layouts can differ by an order of magnitude. We introduce a circuit score used for ranking that is parameterized in terms of a physics-based, phenomenological error model whose parameters are fit by training a ranking-loss function over a measured dataset. The dataset consists of quantum circuits exhibiting a diversity of structures and executed on IBM hardware, allowing the model to incorporate the contextual nature of real device noise and errors without the need to perform an exponentially costly tomographic protocol. We perform model training and execution on the 16-qubit ibmq_guadalupe device and compare our method to two common approaches: random layout selection and a publicly available baseline called Mapomatic. Our model consistently outperforms both approaches, predicting layouts that exhibit lower noise and higher performance. In particular, we find that our best model leads to a $1.8\times$ reduction in selection error when compared to the baseline approach and a $3.2\times$ reduction when compared to random selection. Beyond delivering a new form of predictive quantum characterization, verification, and validation, our results reveal the specific way in which context-dependent and coherent gate errors appear to dominate the divergence from performance estimates extrapolated from simple proxy measures.
Authors: Rachid Karami, Chakshu Moar, Sheng-Chun Kao, Hyoukjun Kwon
Abstract: Machine Learning (ML) operators are the building blocks to design ML models with various target applications. GEneral Matrix Multiplication (GEMM) operators are the backbone of ML models. They are notorious for being computationally expensive requiring billions of multiply-and-accumulate. Therefore, significant effort has been put to study and optimize the GEMM operators in order to speed up the execution of ML models. GPUs and accelerators are widely deployed to accelerate ML workloads by optimizing the execution of GEMM operators. Nonetheless, the performance of NonGEMM operators have not been studied as thoroughly as GEMMs. Therefore, this paper describes \bench, a benchmark to study NonGEMM operators. We first construct \bench using popular ML workloads from different domains, then perform case studies on various grade GPU platforms to analyze the behavior of NonGEMM operators in GPU accelerated systems. Finally, we present some key takeaways to bridge the gap between GEMM and NonGEMM operators and to offer the community with potential new optimization directions.
Authors: Andre Kassis, Urs Hengartner
Abstract: Reports regarding the misuse of Generative AI (GenAI) to create deepfakes are frequent. Defensive watermarking enables GenAI providers to hide fingerprints in their images and use them later for deepfake detection. Yet, its potential has not been fully explored. We present UnMarker -- the first practical universal attack on defensive watermarking. Unlike existing attacks, UnMarker requires no detector feedback, no unrealistic knowledge of the watermarking scheme or similar models, and no advanced denoising pipelines that may not be available. Instead, being the product of an in-depth analysis of the watermarking paradigm revealing that robust schemes must construct their watermarks in the spectral amplitudes, UnMarker employs two novel adversarial optimizations to disrupt the spectra of watermarked images, erasing the watermarks. Evaluations against SOTA schemes prove UnMarker's effectiveness. It not only defeats traditional schemes while retaining superior quality compared to existing attacks but also breaks semantic watermarks that alter an image's structure, reducing the best detection rate to $43\%$ and rendering them useless. To our knowledge, UnMarker is the first practical attack on semantic watermarks, which have been deemed the future of defensive watermarking. Our findings show that defensive watermarking is not a viable defense against deepfakes, and we urge the community to explore alternatives.
Authors: Cristian Bodnar, Wessel P. Bruinsma, Ana Lucic, Megan Stanley, Anna Vaughan, Johannes Brandstetter, Patrick Garvan, Maik Riechert, Jonathan A. Weyn, Haiyu Dong, Jayesh K. Gupta, Kit Thambiratnam, Alexander T. Archibald, Chun-Chieh Wu, Elizabeth Heider, Max Welling, Richard E. Turner, Paris Perdikaris
Abstract: Reliable forecasts of the Earth system are crucial for human progress and safety from natural disasters. Artificial intelligence offers substantial potential to improve prediction accuracy and computational efficiency in this field, however this remains underexplored in many domains. Here we introduce Aurora, a large-scale foundation model for the Earth system trained on over a million hours of diverse data. Aurora outperforms operational forecasts for air quality, ocean waves, tropical cyclone tracks, and high-resolution weather forecasting at orders of magnitude smaller computational expense than dedicated existing systems. With the ability to fine-tune Aurora to diverse application domains at only modest computational cost, Aurora represents significant progress in making actionable Earth system predictions accessible to anyone.
Authors: Xinmeng Huang, Shuo Li, Edgar Dobriban, Osbert Bastani, Hamed Hassani, Dongsheng Ding
Abstract: The growing safety concerns surrounding large language models raise an urgent need to align them with diverse human preferences to simultaneously enhance their helpfulness and safety. A promising approach is to enforce safety constraints through Reinforcement Learning from Human Feedback (RLHF). For such constrained RLHF, typical Lagrangian-based primal-dual policy optimization methods are computationally expensive and often unstable. This paper presents a perspective of dualization that reduces constrained alignment to an equivalent unconstrained alignment problem. We do so by pre-optimizing a smooth and convex dual function that has a closed form. This shortcut eliminates the need for cumbersome primal-dual policy iterations, greatly reducing the computational burden and improving training stability. Our strategy leads to two practical algorithms in model-based and preference-based settings (MoCAN and PeCAN, respectively). A broad range of experiments demonstrate the effectiveness and merits of our algorithms.
Authors: Amil Dravid, Yossi Gandelsman, Kuan-Chieh Wang, Rameen Abdal, Gordon Wetzstein, Alexei A. Efros, Kfir Aberman
Abstract: We investigate the space of weights spanned by a large collection of customized diffusion models. We populate this space by creating a dataset of over 60,000 models, each of which is a base model fine-tuned to insert a different person's visual identity. We model the underlying manifold of these weights as a subspace, which we term weights2weights. We demonstrate three immediate applications of this space that result in new diffusion models -- sampling, editing, and inversion. First, sampling a set of weights from this space results in a new model encoding a novel identity. Next, we find linear directions in this space corresponding to semantic edits of the identity (e.g., adding a beard), resulting in a new model with the original identity edited. Finally, we show that inverting a single image into this space encodes a realistic identity into a model, even if the input image is out of distribution (e.g., a painting). We further find that these linear properties of the diffusion model weight space extend to other visual concepts. Our results indicate that the weight space of fine-tuned diffusion models can behave as an interpretable meta-latent space producing new models.
Authors: Gabriel Sarch, Lawrence Jang, Michael J. Tarr, William W. Cohen, Kenneth Marino, Katerina Fragkiadaki
Abstract: Large-scale generative language and vision-language models (LLMs and VLMs) excel in few-shot in-context learning for decision making and instruction following. However, they require high-quality exemplar demonstrations in their context window. In this work, we ask: Can LLMs and VLMs generate their own examples from generic, sub-optimal demonstrations? We propose In-Context Abstraction Learning (ICAL), a method that builds a memory of multimodal experience from sub-optimal demonstrations and human feedback. Given a task demonstration that may contain inefficiencies or mistakes, a VLM abstracts the trajectory into a generalized program of thoughts by correcting inefficient actions and annotating cognitive abstractions: causal relationships, object state changes, temporal subgoals, and task-relevant visual elements. These programs of thought are iteratively improved through human feedback while the agent executes the trajectory in a similar environment. The resulting examples significantly improve decision-making in retrieval-augmented LLM and VLM agents. Moreover, as the agent's library of examples grows, it becomes more efficient, relying less on human feedback and requiring fewer environment interactions per demonstration. Our ICAL agent surpasses the SOTA in dialogue-based instruction following in TEACh, multimodal web agents in VisualWebArena, and action anticipation in Ego4D. In TEACh, we achieve a 12.6% improvement in goal-condition success. In VisualWebArena, our task success rate improves over few-shot GPT4V. In Ego4D action forecasting, we improve over few-shot GPT-4V and remain competitive with supervised models. We show finetuning our retrieval-augmented in-context agent yields additional improvements. Our approach significantly reduces reliance on manual prompt engineering and consistently outperforms in-context learning from action plans that lack such programs of thought.
Authors: Yifan Yang, Kai Zhen, Ershad Banijamal, Athanasios Mouchtaris, Zheng Zhang
Abstract: Fine-tuning large language models (LLMs) has achieved remarkable performance across various natural language processing tasks, yet it demands more and more memory as model sizes keep growing. To address this issue, the recently proposed Memory-efficient Zeroth-order (MeZO) methods attempt to fine-tune LLMs using only forward passes, thereby avoiding the need for a backpropagation graph. However, significant performance drops and a high risk of divergence have limited their widespread adoption. In this paper, we propose the Adaptive Zeroth-order Tensor-Train Adaption (AdaZeta) framework, specifically designed to improve the performance and convergence of the ZO methods. To enhance dimension-dependent ZO estimation accuracy, we introduce a fast-forward, low-parameter tensorized adapter. To tackle the frequently observed divergence issue in large-scale ZO fine-tuning tasks, we propose an adaptive query number schedule that guarantees convergence. Detailed theoretical analysis and extensive experimental results on Roberta-Large and Llama-2-7B models substantiate the efficacy of our AdaZeta framework in terms of accuracy, memory efficiency, and convergence speed.
Authors: Nikolaos Dionelis, Nicolas Longepe
Abstract: Confidence assessments of semantic segmentation algorithms are important. Ideally, deep learning models should have the ability to predict in advance whether their output is likely to be incorrect. Assessing the confidence levels of model predictions in Earth Observation (EO) classification is essential, as it can enhance semantic segmentation performance and help prevent further exploitation of the results in case of erroneous prediction. The model we developed, Confidence Assessment for enhanced Semantic segmentation (CAS), evaluates confidence at both the segment and pixel levels, providing both labels and confidence scores as output. Our model, CAS, identifies segments with incorrect predicted labels using the proposed combined confidence metric, refines the model, and enhances its performance. This work has significant applications, particularly in evaluating EO Foundation Models on semantic segmentation downstream tasks, such as land cover classification using Sentinel-2 satellite data. The evaluation results show that this strategy is effective and that the proposed model CAS outperforms other baseline models.
Authors: Prasoon Raghuwanshi, Onel Luis Alcaraz L\'opez, Neelesh B. Mehta, Hirley Alves, Matti Latva-aho
Abstract: Efficient Random Access (RA) is critical for enabling reliable communication in Industrial Internet of Things (IIoT) networks. Herein, we propose a deep reinforcement learning based distributed RA scheme, entitled Neural Network-Based Bandit (NNBB), for the IIoT alarm scenario. In such a scenario, the devices may detect a common critical event, and the goal is to ensure the alarm information is delivered successfully from at least one device. The proposed NNBB scheme is implemented at each device, where it trains itself online and establishes implicit inter-device coordination to achieve the common goal. Devices can transmit simultaneously on multiple orthogonal channels and each possible transmission pattern constitutes a possible action for the NNBB, which uses a deep neural network to determine the action. Our simulation results show that as the number of devices in the network increases, so does the performance gain of the NNBB compared to the Multi-Armed Bandit (MAB) RA benchmark. For instance, NNBB experiences a 7% success rate drop when there are four channels and the number of devices increases from 10 to 60, while MAB faces a 25% drop.
Authors: Yun Qu, Boyuan Wang, Jianzhun Shao, Yuhang Jiang, Chen Chen, Zhenbin Ye, Lin Liu, Junfeng Yang, Lin Lai, Hongyang Qin, Minwen Deng, Juchao Zhuo, Deheng Ye, Qiang Fu, Wei Yang, Guang Yang, Lanxiao Huang, Xiangyang Ji
Abstract: The advancement of Offline Reinforcement Learning (RL) and Offline Multi-Agent Reinforcement Learning (MARL) critically depends on the availability of high-quality, pre-collected offline datasets that represent real-world complexities and practical applications. However, existing datasets often fall short in their simplicity and lack of realism. To address this gap, we propose Hokoff, a comprehensive set of pre-collected datasets that covers both offline RL and offline MARL, accompanied by a robust framework, to facilitate further research. This data is derived from Honor of Kings, a recognized Multiplayer Online Battle Arena (MOBA) game known for its intricate nature, closely resembling real-life situations. Utilizing this framework, we benchmark a variety of offline RL and offline MARL algorithms. We also introduce a novel baseline algorithm tailored for the inherent hierarchical action space of the game. We reveal the incompetency of current offline RL approaches in handling task complexity, generalization and multi-task learning.
Authors: Tongzhou Wang, Yihan Lin
Abstract: CycleGAN provides a framework to train image-to-image translation with unpaired datasets using cycle consistency loss [4]. While results are great in many applications, the pixel level cycle consistency can potentially be problematic and causes unrealistic images in certain cases. In this project, we propose three simple modifications to cycle consistency, and show that such an approach achieves better results with fewer artifacts.
Authors: Hiroyasu Inoue, Yasuyuki Todo
Abstract: This study examines how positive and negative news about firms affects their stock prices and, moreover, how it affects stock prices of the firms' suppliers and clients, using a large sample of publicly listed firms around the world and another of Japanese listed firms. The level of positiveness and negativeness of each news article is determined by FinBERT, a natural language processing model fine-tuned specifically for financial information. Supply chains of firms across the world are identified mostly by financial statements, while those of Japanese firms are taken from large-scale firm-level surveys. We find that positive news increases the change rate of stock prices of firms mentioned in the news before its disclosure, most likely because of diffusion of information through private channels. Positive news also raises stock prices of the firms' suppliers and clients before its disclosure, confirming propagation of market values through supply chains. In addition, we generally find a larger post-news effect on stock prices of the mentioned firms and their suppliers and clients than the pre-news effect. The positive difference between the post- and pre-news effects can be considered as the net effect of the disclosure of positive news, controlling for information diffusion through private channels. However, the post-news effect on suppliers and clients in Japan is smaller than the pre-news effect, which is the opposite result to non-domestic firms from around the world.
Authors: Kosuke Imai, Kentaro Nakamura
Abstract: In this paper, we demonstrate how to enhance the validity of causal inference with unstructured high-dimensional treatments like texts, by leveraging the power of generative Artificial Intelligence. Specifically, we propose to use a deep generative model such as large language models (LLMs) to efficiently generate treatments and use their internal representation for subsequent causal effect estimation. We show that the knowledge of this true internal representation helps disentangle the treatment features of interest, such as specific sentiments and certain topics, from other possibly unknown confounding features. Unlike the existing methods, our proposed approach eliminates the need to learn causal representation from the data and hence produces more accurate and efficient estimates. We formally establish the conditions required for the nonparametric identification of the average treatment effect, propose an estimation strategy that avoids the violation of the overlap assumption, and derive the asymptotic properties of the proposed estimator through the application of double machine learning. Finally, using an instrumental variables approach, we extend the proposed methodology to the settings, in which the treatment feature is based on human perception rather than is assumed to be fixed given the treatment object. The proposed methodology is also applicable to text reuse where an LLM is used to regenerate the existing texts. We conduct simulation and empirical studies, using the generated text data from an open-source LLM, Llama 3, to illustrate the advantages of our estimator over the state-of-the-art causal representation learning algorithms.
Authors: Cristian Daniel P\u{a}duraru, Antonio B\u{a}rb\u{a}lau, Radu Filipescu, Andrei Liviu Nicolicioiu, Elena Burceanu
Abstract: An important goal of ML research is to identify and mitigate unwanted biases intrinsic to datasets and already incorporated into pre-trained models. Previous approaches have identified biases using highly curated validation subsets, that require human knowledge to create in the first place. This limits the ability to automate the discovery of unknown biases in new datasets. We solve this by using interpretable vision-language models, combined with a filtration method using LLMs and known concept hierarchies. More exactly, for a dataset, we use pre-trained CLIP models that have an associated embedding for each class and see how it drifts through learning towards embeddings that disclose hidden biases. We call this approach ConceptDrift and show that it can be scaled to automatically identify biases in datasets like ImageNet without human prior knowledge. We propose two bias identification evaluation protocols to fill the gap in the previous work and show that our method significantly improves over SoTA methods, both using our protocol and classical evaluations. Alongside validating the identified biases, we also show that they can be leveraged to improve the performance of different methods. Our method is not bounded to a single modality, and we empirically validate it both on image (Waterbirds, CelebA, ImageNet), and text datasets (CivilComments).
Authors: Haolin Chen, Yihao Feng, Zuxin Liu, Weiran Yao, Akshara Prabhakar, Shelby Heinecke, Ricky Ho, Phil Mui, Silvio Savarese, Caiming Xiong, Huan Wang
Abstract: Large language models (LLMs) have shown impressive capabilities, but still struggle with complex reasoning tasks requiring multiple steps. While prompt-based methods like Chain-of-Thought (CoT) can improve LLM reasoning at inference time, optimizing reasoning capabilities during training remains challenging. We introduce LaTent Reasoning Optimization (LaTRO), a principled framework that formulates reasoning as sampling from a latent distribution and optimizes it via variational approaches. LaTRO enables LLMs to concurrently improve both their reasoning process and ability to evaluate reasoning quality, without requiring external feedback or reward models. We validate LaTRO through experiments on GSM8K and ARC-Challenge datasets using multiple model architectures. On GSM8K, LaTRO improves zero-shot accuracy by an average of 12.5% over base models and 9.6% over supervised fine-tuning across Phi-3.5-mini, Mistral-7B, and Llama-3.1-8B. Our findings suggest that pre-trained LLMs possess latent reasoning capabilities that can be unlocked and enhanced through our proposed optimization approach in a self-improvement manner. The code of LaTRO is available at \url{https://github.com/SalesforceAIResearch/LaTRO}.
Authors: Guixian Xu, Jinglai Li, Junqi Tang
Abstract: Equivariant Imaging (EI) regularization has become the de-facto technique for unsupervised training of deep imaging networks, without any need of ground-truth data. Observing that the EI-based unsupervised training paradigm currently has significant computational redundancy leading to inefficiency in high-dimensional applications, we propose a sketched EI regularization which leverages the randomized sketching techniques for acceleration. We then extend our sketched EI regularization to develop an accelerated deep internal learning framework -- Sketched Equivariant Deep Image Prior (Sk-EI-DIP), which can be efficiently applied for single-image and task-adapted reconstruction. Additionally, for network adaptation tasks, we propose a parameter-efficient approach for accelerating both EI-DIP and Sk-EI-DIP via optimizing only the normalization layers. Our numerical study on X-ray CT image reconstruction tasks demonstrate that our approach can achieve order-of-magnitude computational acceleration over standard EI-based counterpart in single-input setting, and network adaptation at test time.
Authors: Jongmin Mun, Seungwoo Kwak, Ilmun Kim
Abstract: We explore the trade-off between privacy and statistical utility in private two-sample testing under local differential privacy (LDP) for both multinomial and continuous data. We begin by addressing the multinomial case, where we introduce private permutation tests using practical privacy mechanisms such as Laplace, discrete Laplace, and Google's RAPPOR. We then extend our multinomial approach to continuous data via binning and study its uniform separation rates under LDP over H\"older and Besov smoothness classes. The proposed tests for both discrete and continuous cases rigorously control the type I error for any finite sample size, strictly adhere to LDP constraints, and achieve minimax separation rates under LDP. The attained minimax rates reveal inherent privacy-utility trade-offs that are unavoidable in private testing. To address scenarios with unknown smoothness parameters in density testing, we propose an adaptive test based on a Bonferroni-type approach that ensures robust performance without prior knowledge of the smoothness parameters. We validate our theoretical findings with extensive numerical experiments and demonstrate the practical relevance and effectiveness of our proposed methods.
Authors: Maraz Mia, Darius Derakhshan, Mir Mehedi A. Pritom
Abstract: Phishing has been a prevalent cyber threat that manipulates users into revealing sensitive private information through deceptive tactics, designed to masquerade as trustworthy entities. Over the years, proactively detection of phishing URLs (or websites) has been established as an widely-accepted defense approach. In literature, we often find supervised Machine Learning (ML) models with highly competitive performance for detecting phishing websites based on the extracted features from both phishing and benign (i.e., legitimate) websites. However, it is still unclear if these features or indicators are dependent on a particular dataset or they are generalized for overall phishing detection. In this paper, we delve deeper into this issue by analyzing two publicly available phishing URL datasets, where each dataset has its own set of unique and overlapping features related to URL string and website contents. We want to investigate if overlapping features are similar in nature across datasets and how does the model perform when trained on one dataset and tested on the other. We conduct practical experiments and leverage explainable AI (XAI) methods such as SHAP plots to provide insights into different features' contributions in case of phishing detection to answer our primary question, "Can features for phishing URL detection be trusted across diverse dataset?". Our case study experiment results show that features for phishing URL detection can often be dataset-dependent and thus may not be trusted across different datasets even though they share same set of feature behaviors.
Authors: Mark Sch\"one, Yash Bhisikar, Karan Bania, Khaleelulla Khan Nazeer, Christian Mayr, Anand Subramoney, David Kappel
Abstract: Handling sparse and unstructured geometric data, such as point clouds or event-based vision, is a pressing challenge in the field of machine vision. Recently, sequence models such as Transformers and state-space models entered the domain of geometric data. These methods require specialized preprocessing to create a sequential view of a set of points. Furthermore, prior works involving sequence models iterate geometric data with either uniform or learned step sizes, implicitly relying on the model to infer the underlying geometric structure. In this work, we propose to encode geometric structure explicitly into the parameterization of a state-space model. State-space models are based on linear dynamics governed by a one-dimensional variable such as time or a spatial coordinate. We exploit this dynamic variable to inject relative differences of coordinates into the step size of the state-space model. The resulting geometric operation computes interactions between all pairs of N points in O(N) steps. Our model deploys the Mamba selective state-space model with a modified CUDA kernel to efficiently map sparse geometric data to modern hardware. The resulting sequence model, which we call STREAM, achieves competitive results on a range of benchmarks from point-cloud classification to event-based vision and audio classification. STREAM demonstrates a powerful inductive bias for sparse geometric data by improving the PointMamba baseline when trained from scratch on the ModelNet40 and ScanObjectNN point cloud analysis datasets. It further achieves, for the first time, 100% test accuracy on all 11 classes of the DVS128 Gestures dataset.
Authors: Cl\'ement Bonnet, Ariel N. Lee, Franck Wertel, Antoine Tamano, Tanguy Cizain, Pablo Ducru
Abstract: In the last two years, text-to-image diffusion models have become extremely popular. As their quality and usage increase, a major concern has been the need for better output control. In addition to prompt engineering, one effective method to improve the controllability of diffusion models has been to condition them on additional modalities such as image style, depth map, or keypoints. This forms the basis of ControlNets or Adapters. When attempting to apply these methods to control human poses in outputs of text-to-image diffusion models, two main challenges have arisen. The first challenge is generating poses following a wide range of semantic text descriptions, for which previous methods involved searching for a pose within a dataset of (caption, pose) pairs. The second challenge is conditioning image generation on a specified pose while keeping both high aesthetic and high pose fidelity. In this article, we fix these two main issues by introducing a text-to-pose (T2P) generative model alongside a new sampling algorithm, and a new pose adapter that incorporates more pose keypoints for higher pose fidelity. Together, these two new state-of-the-art models enable, for the first time, a generative text-to-pose-to-image framework for higher pose control in diffusion models. We release all models and the code used for the experiments at https://github.com/clement-bonnet/text-to-pose.
Authors: John D. Hastings, Sherri Weitl-Harms, Joseph Doty, Zachary J. Myers, Warren Thompson
Abstract: This research explores the application of large language models (LLMs) to generate synthetic datasets for Product Desirability Toolkit (PDT) testing, a key component in evaluating user sentiment and product experience. Utilizing gpt-4o-mini, a cost-effective alternative to larger commercial LLMs, three methods, Word+Review, Review+Word, and Supply-Word, were each used to synthesize 1000 product reviews. The generated datasets were assessed for sentiment alignment, textual diversity, and data generation cost. Results demonstrated high sentiment alignment across all methods, with Pearson correlations ranging from 0.93 to 0.97. Supply-Word exhibited the highest diversity and coverage of PDT terms, although with increased generation costs. Despite minor biases toward positive sentiments, in situations with limited test data, LLM-generated synthetic data offers significant advantages, including scalability, cost savings, and flexibility in dataset production.
Authors: An Wang, Li Zou
Abstract: We propose a unified framework based on persistent homology (PH) to characterize both local and global structures in disordered systems. It can simultaneously generate local and global descriptors using the same algorithm and data structure, and has shown to be highly effective and interpretable in predicting particle rearrangements and classifying global phases. Based on this framework, we define a non-parametric metric, the Separation Index (SI), which not only outperforms traditional bond-orientational order parameters in phase classification tasks but also establishes a connection between particle environments and the global phase structure. Our methods provide an effective framework for understanding and analyzing the properties of disordered materials, with broad potential applications in materials science and even wider studies of complex systems.