id
stringlengths
9
16
title
stringlengths
4
278
categories
stringlengths
5
104
abstract
stringlengths
6
4.09k
2502.08649
Principles for Open Data Curation: A Case Study with the New York City 311 Service Request Data
cs.DB cs.CY stat.ME
In the early 21st century, the open data movement began to transform societies and governments by promoting transparency, innovation, and public engagement. The City of New York (NYC) has been at the forefront of this movement since the enactment of the Open Data Law in 2012, creating the NYC Open Data portal. The portal currently hosts 2,700 datasets, serving as a crucial resource for research across various domains, including health, urban development, and transportation. However, the effective use of open data relies heavily on data quality and usability, challenges that remain insufficiently addressed in the literature. This paper examines these challenges via a case study of the NYC 311 Service Request dataset, identifying key issues in data validity, consistency, and curation efficiency. We propose a set of data curation principles, tailored for government-released open data, to address these challenges. Our findings highlight the importance of harmonized field definitions, streamlined storage, and automated quality checks, offering practical guidelines for improving the reliability and utility of open datasets.
2502.08651
Democratizing AI Governance: Balancing Expertise and Public Participation
cs.CY cs.LG
The development and deployment of artificial intelligence (AI) systems, with their profound societal impacts, raise critical challenges for governance. Historically, technological innovations have been governed by concentrated expertise with limited public input. However, AI's pervasive influence across domains such as healthcare, employment, and justice necessitates inclusive governance approaches. This article explores the tension between expert-led oversight and democratic participation, analyzing models of participatory and deliberative democracy. Using case studies from France and Brazil, we highlight how inclusive frameworks can bridge the gap between technical complexity and public accountability. Recommendations are provided for integrating these approaches into a balanced governance model tailored to the European Union, emphasizing transparency, diversity, and adaptive regulation to ensure that AI governance reflects societal values while maintaining technical rigor. This analysis underscores the importance of hybrid frameworks that unite expertise and public voice in shaping the future of AI policy.
2502.08652
LegalScore: Development of a Benchmark for Evaluating AI Models in Legal Career Exams in Brazil
cs.CY cs.AI
This research introduces LegalScore, a specialized index for assessing how generative artificial intelligence models perform in a selected range of career exams that require a legal background in Brazil. The index evaluates fourteen different types of artificial intelligence models' performance, from proprietary to open-source models, in answering objective questions applied to these exams. The research uncovers the response of the models when applying English-trained large language models to Brazilian legal contexts, leading us to reflect on the importance and the need for Brazil-specific training data in generative artificial intelligence models. Performance analysis shows that while proprietary and most known models achieved better results overall, local and smaller models indicated promising performances due to their Brazilian context alignment in training. By establishing an evaluation framework with metrics including accuracy, confidence intervals, and normalized scoring, LegalScore enables systematic assessment of artificial intelligence performance in legal examinations in Brazil. While the study demonstrates artificial intelligence's potential value for exam preparation and question development, it concludes that significant improvements are needed before AI can match human performance in advanced legal assessments. The benchmark creates a foundation for continued research, highlighting the importance of local adaptation in artificial intelligence development.
2502.08655
Personalizing Education through an Adaptive LMS with Integrated LLMs
cs.AI
The widespread adoption of large language models (LLMs) marks a transformative era in technology, especially within the educational sector. This paper explores the integration of LLMs within learning management systems (LMSs) to develop an adaptive learning management system (ALMS) personalized for individual learners across various educational stages. Traditional LMSs, while facilitating the distribution of educational materials, fall short in addressing the nuanced needs of diverse student populations, particularly in settings with limited instructor availability. Our proposed system leverages the flexibility of AI to provide a customizable learning environment that adjusts to each user's evolving needs. By integrating a suite of general-purpose and domain-specific LLMs, this system aims to minimize common issues such as factual inaccuracies and outdated information, characteristic of general LLMs like OpenAI's ChatGPT. This paper details the development of an ALMS that not only addresses privacy concerns and the limitations of existing educational tools but also enhances the learning experience by maintaining engagement through personalized educational content.
2502.08657
Refining Positive and Toxic Samples for Dual Safety Self-Alignment of LLMs with Minimal Human Interventions
cs.CL cs.AI
Recent AI agents, such as ChatGPT and LLaMA, primarily rely on instruction tuning and reinforcement learning to calibrate the output of large language models (LLMs) with human intentions, ensuring the outputs are harmless and helpful. Existing methods heavily depend on the manual annotation of high-quality positive samples, while contending with issues such as noisy labels and minimal distinctions between preferred and dispreferred response data. However, readily available toxic samples with clear safety distinctions are often filtered out, removing valuable negative references that could aid LLMs in safety alignment. In response, we propose PT-ALIGN, a novel safety self-alignment approach that minimizes human supervision by automatically refining positive and toxic samples and performing fine-grained dual instruction tuning. Positive samples are harmless responses, while toxic samples deliberately contain extremely harmful content, serving as a new supervisory signals. Specifically, we utilize LLM itself to iteratively generate and refine training instances by only exploring fewer than 50 human annotations. We then employ two losses, i.e., maximum likelihood estimation (MLE) and fine-grained unlikelihood training (UT), to jointly learn to enhance the LLM's safety. The MLE loss encourages an LLM to maximize the generation of harmless content based on positive samples. Conversely, the fine-grained UT loss guides the LLM to minimize the output of harmful words based on negative samples at the token-level, thereby guiding the model to decouple safety from effectiveness, directing it toward safer fine-tuning objectives, and increasing the likelihood of generating helpful and reliable content. Experiments on 9 popular open-source LLMs demonstrate the effectiveness of our PT-ALIGN for safety alignment, while maintaining comparable levels of helpfulness and usefulness.
2502.08658
Analyzable Parameters Dominated Vehicle Platoon Dynamics Modeling and Analysis: A Physics-Encoded Deep Learning Approach
cs.RO cs.AI
Recently, artificial intelligence (AI)-enabled nonlinear vehicle platoon dynamics modeling plays a crucial role in predicting and optimizing the interactions between vehicles. Existing efforts lack the extraction and capture of vehicle behavior interaction features at the platoon scale. More importantly, maintaining high modeling accuracy without losing physical analyzability remains to be solved. To this end, this paper proposes a novel physics-encoded deep learning network, named PeMTFLN, to model the nonlinear vehicle platoon dynamics. Specifically, an analyzable parameters encoded computational graph (APeCG) is designed to guide the platoon to respond to the driving behavior of the lead vehicle while ensuring local stability. Besides, a multi-scale trajectory feature learning network (MTFLN) is constructed to capture platoon following patterns and infer the physical parameters required for APeCG from trajectory data. The human-driven vehicle trajectory datasets (HIGHSIM) were used to train the proposed PeMTFLN. The trajectories prediction experiments show that PeMTFLN exhibits superior compared to the baseline models in terms of predictive accuracy in speed and gap. The stability analysis result shows that the physical parameters in APeCG is able to reproduce the platoon stability in real-world condition. In simulation experiments, PeMTFLN performs low inference error in platoon trajectories generation. Moreover, PeMTFLN also accurately reproduces ground-truth safety statistics. The code of proposed PeMTFLN is open source.
2502.08659
Deployment-friendly Lane-changing Intention Prediction Powered by Brain-inspired Spiking Neural Networks
cs.RO
Accurate and real-time prediction of surrounding vehicles' lane-changing intentions is a critical challenge in deploying safe and efficient autonomous driving systems in open-world scenarios. Existing high-performing methods remain hard to deploy due to their high computational cost, long training times, and excessive memory requirements. Here, we propose an efficient lane-changing intention prediction approach based on brain-inspired Spiking Neural Networks (SNN). By leveraging the event-driven nature of SNN, the proposed approach enables us to encode the vehicle's states in a more efficient manner. Comparison experiments conducted on HighD and NGSIM datasets demonstrate that our method significantly improves training efficiency and reduces deployment costs while maintaining comparable prediction accuracy. Particularly, compared to the baseline, our approach reduces training time by 75% and memory usage by 99.9%. These results validate the efficiency and reliability of our method in lane-changing predictions, highlighting its potential for safe and efficient autonomous driving systems while offering significant advantages in deployment, including reduced training time, lower memory usage, and faster inference.
2502.08660
Semantic Role Labeling: A Systematical Survey
cs.CL
Semantic role labeling (SRL) is a central natural language processing (NLP) task aiming to understand the semantic roles within texts, facilitating a wide range of downstream applications. While SRL has garnered extensive and enduring research, there is currently a lack of a comprehensive survey that thoroughly organizes and synthesizes the field. This paper aims to review the entire research trajectory of the SRL community over the past two decades. We begin by providing a complete definition of SRL. To offer a comprehensive taxonomy, we categorize SRL methodologies into four key perspectives: model architectures, syntax feature modeling, application scenarios, and multi-modal extensions. Further, we discuss SRL benchmarks, evaluation metrics, and paradigm modeling approaches, while also exploring practical applications across various domains. Finally, we analyze future research directions in SRL, addressing the evolving role of SRL in the age of large language models (LLMs) and its potential impact on the broader NLP landscape. We maintain a public repository and consistently update related resources at: https://github.com/DreamH1gh/Awesome-SRL
2502.08661
Few-shot LLM Synthetic Data with Distribution Matching
cs.CL cs.AI
As large language models (LLMs) advance, their ability to perform in-context learning and few-shot language generation has improved significantly. This has spurred using LLMs to produce high-quality synthetic data to enhance the performance of smaller models like online retrievers or weak LLMs. However, LLM-generated synthetic data often differs from the real data in key language attributes (e.g., styles, tones, content proportions, etc.). As a result, mixing these synthetic data directly with real data may distort the original data distribution, potentially hindering performance improvements. To solve this, we introduce SynAlign: a synthetic data generation and filtering framework based on key attribute distribution matching. Before generation, SynAlign employs an uncertainty tracker surrogated by the Gaussian Process model to iteratively select data clusters distinct from selected ones as demonstrations for new data synthesis, facilitating the efficient exploration diversity of the real data. Then, a latent attribute reasoning method is employed: the LLM summarizes linguistic attributes of demonstrations and then synthesizes new data based on them. This approach facilitates synthesizing diverse data with linguistic attributes that appear in real data.After generation, the Maximum Mean Discrepancy is used as the objective function to learn the sampling weight of each synthetic data, ensuring distribution matching with the real data. Our experiments on multiple text prediction tasks show significant performance improvements. We also conducted an online A/B test on an online retriever to demonstrate SynAlign's effectiveness.
2502.08662
RoToR: Towards More Reliable Responses for Order-Invariant Inputs
cs.CL cs.AI
Mitigating positional bias of language models (LMs) for listwise inputs is a well-known and important problem (e.g., lost-in-the-middle). While zero-shot order-invariant LMs have been proposed to solve this issue, their success on practical listwise problems has been limited. In this work, as a first contribution, we identify and overcome two limitations to make zero-shot invariant LMs more practical: (1) training and inference distribution mismatch arising from modifying positional ID assignments to enforce invariance, and (2) failure to adapt to a mixture of order-invariant and sensitive inputs in practical listwise problems. To overcome, we propose (1) RoToR, a zero-shot invariant LM for genuinely order-invariant inputs with minimal modifications of positional IDs, and (2) Selective Routing, an adaptive framework that handles both order-invariant and order-sensitive inputs in listwise tasks. On the Lost in the middle (LitM), Knowledge Graph Question Answering (KGQA), and MMLU benchmarks, we show that RoToR with Selective Routing can effectively handle practical listwise input tasks in a zero-shot manner.
2502.08663
Hallucination Detection: A Probabilistic Framework Using Embeddings Distance Analysis
cs.CL cs.AI cs.CY
Hallucinations are one of the major issues affecting LLMs, hindering their wide adoption in production systems. While current research solutions for detecting hallucinations are mainly based on heuristics, in this paper we introduce a mathematically sound methodology to reason about hallucination, and leverage it to build a tool to detect hallucinations. To the best of our knowledge, we are the first to show that hallucinated content has structural differences with respect to correct content. To prove this result, we resort to the Minkowski distances in the embedding space. Our findings demonstrate statistically significant differences in the embedding distance distributions, that are also scale free -- they qualitatively hold regardless of the distance norm used and the number of keywords, questions, or responses. We leverage these structural differences to develop a tool to detect hallucinated responses, achieving an accuracy of 66\% for a specific configuration of system parameters -- comparable with the best results in the field. In conclusion, the suggested methodology is promising and novel, possibly paving the way for further research in the domain, also along the directions highlighted in our future work.
2502.08664
Motion Forecasting for Autonomous Vehicles: A Survey
cs.RO cs.AI
In recent years, the field of autonomous driving has attracted increasingly significant public interest. Accurately forecasting the future behavior of various traffic participants is essential for the decision-making of Autonomous Vehicles (AVs). In this paper, we focus on both scenario-based and perception-based motion forecasting for AVs. We propose a formal problem formulation for motion forecasting and summarize the main challenges confronting this area of research. We also detail representative datasets and evaluation metrics pertinent to this field. Furthermore, this study classifies recent research into two main categories: supervised learning and self-supervised learning, reflecting the evolving paradigms in both scenario-based and perception-based motion forecasting. In the context of supervised learning, we thoroughly examine and analyze each key element of the methodology. For self-supervised learning, we summarize commonly adopted techniques. The paper concludes and discusses potential research directions, aiming to propel progress in this vital area of AV technology.
2502.08666
Hallucination, Monofacts, and Miscalibration: An Empirical Investigation
cs.CL cs.AI
Recent theoretical work by [Kalai and Vempala 2024] proves that a particular notion of hallucination rate in LLMs must be lower bounded by the training data monofact rate (related to the classical Good-Turing missing mass estimator) minus model miscalibration. Through systematic experiments with n-gram models and in-context learning with LLMs, we empirically investigate and validate this theory by examining how different underlying data distributions affect the monofact rate and a model's tendency to hallucinate. We then vary model miscalibration through controlled upweighting of training samples while holding monofact rates constant, allowing us to isolate miscalibration's reduction effect on hallucination. These findings suggest that both the distribution of fact frequencies in training data and the calibration-hallucination trade-off are inherent to probabilistic language generation. Our results also suggest that current practices of aggressive deduplication in training data may need to be reconsidered, as selective duplication could serve as a principled mechanism for reducing hallucination.
2502.08667
Unpaired Image-to-Image Translation with Content Preserving Perspective: A Review
eess.IV cs.CV
Image-to-image translation (I2I) transforms an image from a source domain to a target domain while preserving source content. Most computer vision applications are in the field of image-to-image translation, such as style transfer, image segmentation, and photo enhancement. The degree of preservation of the content of the source images in the translation process can be different according to the problem and the intended application. From this point of view, in this paper, we divide the different tasks in the field of image-to-image translation into three categories: Fully Content preserving, Partially Content preserving, and Non-Content preserving. We present different tasks, datasets, methods, results of methods for these three categories in this paper. We make a categorization for I2I methods based on the architecture of different models and study each category separately. In addition, we introduce well-known evaluation criteria in the I2I translation field. Specifically, nearly 70 different I2I models were analyzed, and more than 10 quantitative evaluation metrics and 30 distinct tasks and datasets relevant to the I2I translation problem were both introduced and assessed. Translating from simulation to real images could be well viewed as an application of fully content preserving or partially content preserving unsupervised image-to-image translation methods. So, we provide a benchmark for Sim-to-Real translation, which can be used to evaluate different methods. In general, we conclude that because of the different extent of the obligation to preserving content in various applications, it is better to consider this issue in choosing a suitable I2I model for a specific application.
2502.08668
Style Extraction on Text Embeddings Using VAE and Parallel Dataset
cs.CL
This study investigates the stylistic differences among various Bible translations using a Variational Autoencoder (VAE) model. By embedding textual data into high-dimensional vectors, the study aims to detect and analyze stylistic variations between translations, with a specific focus on distinguishing the American Standard Version (ASV) from other translations. The results demonstrate that each translation exhibits a unique stylistic distribution, which can be effectively identified using the VAE model. These findings suggest that the VAE model is proficient in capturing and differentiating textual styles, although it is primarily optimized for distinguishing a single style. The study highlights the model's potential for broader applications in AI-based text generation and stylistic analysis, while also acknowledging the need for further model refinement to address the complexity of multi-dimensional stylistic relationships. Future research could extend this methodology to other text domains, offering deeper insights into the stylistic features embedded within various types of textual data.
2502.08669
Assessing the Impact of the Quality of Textual Data on Feature Representation and Machine Learning Models
cs.CL
Background: Data collected in controlled settings typically results in high-quality datasets. However, in real-world applications, the quality of data collection is often compromised. It is well established that the quality of a dataset significantly impacts the performance of machine learning models. Methods: A rudimentary error rate metric was developed to evaluate textual dataset quality at the token level. Mixtral Large Language Model (LLM) was used to quantify and correct errors in low quality datasets. The study analyzed two healthcare datasets: the high-quality MIMIC-III public hospital dataset and a lower-quality private dataset from Australian aged care homes. Errors were systematically introduced into MIMIC at varying rates, while the ACH dataset quality was improved using the LLM. Results: For the sampled 35,774 and 6,336 patients from the MIMIC and ACH datasets respectively, we used Mixtral to introduce errors in MIMIC and correct errors in ACH. Mixtral correctly detected errors in 63% of progress notes, with 17% containing a single token misclassified due to medical terminology. LLMs demonstrated potential for improving progress note quality by addressing various errors. Under varying error rates, feature representation performance was tolerant to lower error rates (<10%) but declined significantly at higher rates. Conclusions: The study revealed that models performed relatively well on datasets with lower error rates (<10%), but their performance declined significantly as error rates increased (>=10%). Therefore, it is crucial to evaluate the quality of a dataset before utilizing it for machine learning tasks. For datasets with higher error rates, implementing corrective measures is essential to ensure the reliability and effectiveness of machine learning models.
2502.08671
Color Universal Design Neural Network for the Color Vision Deficiencies
eess.IV cs.CV
Information regarding images should be visually understood by anyone, including those with color deficiency. However, such information is not recognizable if the color that seems to be distorted to the color deficiencies meets an adjacent object. The aim of this paper is to propose a color universal design network, called CUD-Net, that generates images that are visually understandable by individuals with color deficiency. CUD-Net is a convolutional deep neural network that can preserve color and distinguish colors for input images by regressing the node point of a piecewise linear function and using a specific filter for each image. To generate CUD images for color deficiencies, we follow a four-step process. First, we refine the CUD dataset based on specific criteria by color experts. Second, we expand the input image information through pre-processing that is specialized for color deficiency vision. Third, we employ a multi-modality fusion architecture to combine features and process the expanded images. Finally, we propose a conjugate loss function based on the composition of the predicted image through the model to address one-to-many problems that arise from the dataset. Our approach is able to produce high-quality CUD images that maintain color and contrast stability. The code for CUD-Net is available on the GitHub repository
2502.08673
High-Throughput SAT Sampling
cs.AI cs.LG
In this work, we present a novel technique for GPU-accelerated Boolean satisfiability (SAT) sampling. Unlike conventional sampling algorithms that directly operate on conjunctive normal form (CNF), our method transforms the logical constraints of SAT problems by factoring their CNF representations into simplified multi-level, multi-output Boolean functions. It then leverages gradient-based optimization to guide the search for a diverse set of valid solutions. Our method operates directly on the circuit structure of refactored SAT instances, reinterpreting the SAT problem as a supervised multi-output regression task. This differentiable technique enables independent bit-wise operations on each tensor element, allowing parallel execution of learning processes. As a result, we achieve GPU-accelerated sampling with significant runtime improvements ranging from $33.6\times$ to $523.6\times$ over state-of-the-art heuristic samplers. We demonstrate the superior performance of our sampling method through an extensive evaluation on $60$ instances from a public domain benchmark suite utilized in previous studies.
2502.08674
COutfitGAN: Learning to Synthesize Compatible Outfits Supervised by Silhouette Masks and Fashion Styles
cs.CV cs.GR cs.MM
How to recommend outfits has gained considerable attention in both academia and industry in recent years. Many studies have been carried out regarding fashion compatibility learning, to determine whether the fashion items in an outfit are compatible or not. These methods mainly focus on evaluating the compatibility of existing outfits and rarely consider applying such knowledge to 'design' new fashion items. We propose the new task of generating complementary and compatible fashion items based on an arbitrary number of given fashion items. In particular, given some fashion items that can make up an outfit, the aim of this paper is to synthesize photo-realistic images of other, complementary, fashion items that are compatible with the given ones. To achieve this, we propose an outfit generation framework, referred to as COutfitGAN, which includes a pyramid style extractor, an outfit generator, a UNet-based real/fake discriminator, and a collocation discriminator. To train and evaluate this framework, we collected a large-scale fashion outfit dataset with over 200K outfits and 800K fashion items from the Internet. Extensive experiments show that COutfitGAN outperforms other baselines in terms of similarity, authenticity, and compatibility measurements.
2502.08676
LIR-LIVO: A Lightweight,Robust LiDAR/Vision/Inertial Odometry with Illumination-Resilient Deep Features
cs.RO cs.CV cs.SY eess.SP eess.SY
In this paper, we propose LIR-LIVO, a lightweight and robust LiDAR-inertial-visual odometry system designed for challenging illumination and degraded environments. The proposed method leverages deep learning-based illumination-resilient features and LiDAR-Inertial-Visual Odometry (LIVO). By incorporating advanced techniques such as uniform depth distribution of features enabled by depth association with LiDAR point clouds and adaptive feature matching utilizing Superpoint and LightGlue, LIR-LIVO achieves state-of-the-art (SOTA) accuracy and robustness with low computational cost. Experiments are conducted on benchmark datasets, including NTU-VIRAL, Hilti'22, and R3LIVE-Dataset. The corresponding results demonstrate that our proposed method outperforms other SOTA methods on both standard and challenging datasets. Particularly, the proposed method demonstrates robust pose estimation under poor ambient lighting conditions in the Hilti'22 dataset. The code of this work is publicly accessible on GitHub to facilitate advancements in the robotics community.
2502.08678
Multispectral Remote Sensing for Weed Detection in West Australian Agricultural Lands
cs.CV eess.IV
The Kondinin region in Western Australia faces significant agricultural challenges due to pervasive weed infestations, causing economic losses and ecological impacts. This study constructs a tailored multispectral remote sensing dataset and an end-to-end framework for weed detection to advance precision agriculture practices. Unmanned aerial vehicles were used to collect raw multispectral data from two experimental areas (E2 and E8) over four years, covering 0.6046 km^{2} and ground truth annotations were created with GPS-enabled vehicles to manually label weeds and crops. The dataset is specifically designed for agricultural applications in Western Australia. We propose an end-to-end framework for weed detection that includes extensive preprocessing steps, such as denoising, radiometric calibration, image alignment, orthorectification, and stitching. The proposed method combines vegetation indices (NDVI, GNDVI, EVI, SAVI, MSAVI) with multispectral channels to form classification features, and employs several deep learning models to identify weeds based on the input features. Among these models, ResNet achieves the highest performance, with a weed detection accuracy of 0.9213, an F1-Score of 0.8735, an mIOU of 0.7888, and an mDC of 0.8865, validating the efficacy of the dataset and the proposed weed detection method.
2502.08679
Deep Learning-Driven Malware Classification with API Call Sequence Analysis and Concept Drift Handling
cs.LG cs.AI cs.CR
Malware classification in dynamic environments presents a significant challenge due to concept drift, where the statistical properties of malware data evolve over time, complicating detection efforts. To address this issue, we propose a deep learning framework enhanced with a genetic algorithm to improve malware classification accuracy and adaptability. Our approach incorporates mutation operations and fitness score evaluations within genetic algorithms to continuously refine the deep learning model, ensuring robustness against evolving malware threats. Experimental results demonstrate that this hybrid method significantly enhances classification performance and adaptability, outperforming traditional static models. Our proposed approach offers a promising solution for real-time malware classification in ever-changing cybersecurity landscapes.
2502.08680
Mathematical Reasoning in Large Language Models: Assessing Logical and Arithmetic Errors across Wide Numerical Ranges
cs.LG cs.AI cs.CL
Mathematical reasoning in Large Language Models (LLMs) is often evaluated using benchmarks with limited numerical ranges, failing to reflect real-world problem-solving across diverse scales. Furthermore, most existing evaluation methods only compare model outputs to ground-truth answers, obscuring insights into reasoning processes. To address these limitations, we introduce GSM-Ranges, a dataset generator derived from GSM8K that systematically perturbs numerical values in math problems to assess model robustness across varying numerical scales. Additionally, we propose a novel grading methodology that distinguishes between logical and non-logical errors, offering a more precise evaluation of reasoning processes beyond computational accuracy. Our experiments with various models reveal a significant increase in logical error rates-up to 14 percentage points-as numerical complexity rises, demonstrating a general weakness in reasoning with out-of-distribution numerical values. Moreover, while models demonstrate high accuracy on standalone arithmetic tasks, their performance deteriorates substantially when computations are embedded within word problems. These findings provide a comprehensive evaluation of LLMs' mathematical reasoning capabilities and inform future research directions for improving numerical generalization in language models.
2502.08681
Centrally Coordinated Multi-Agent Reinforcement Learning for Power Grid Topology Control
cs.MA cs.AI cs.LG
Power grid operation is becoming more complex due to the increase in generation of renewable energy. The recent series of Learning To Run a Power Network (L2RPN) competitions have encouraged the use of artificial agents to assist human dispatchers in operating power grids. However, the combinatorial nature of the action space poses a challenge to both conventional optimizers and learned controllers. Action space factorization, which breaks down decision-making into smaller sub-tasks, is one approach to tackle the curse of dimensionality. In this study, we propose a centrally coordinated multi-agent (CCMA) architecture for action space factorization. In this approach, regional agents propose actions and subsequently a coordinating agent selects the final action. We investigate several implementations of the CCMA architecture, and benchmark in different experimental settings against various L2RPN baseline approaches. The CCMA architecture exhibits higher sample efficiency and superior final performance than the baseline approaches. The results suggest high potential of the CCMA approach for further application in higher-dimensional L2RPN as well as real-world power grid settings.
2502.08682
On the Role of Pre-trained Embeddings in Binary Code Analysis
cs.LG cs.AI
Deep learning has enabled remarkable progress in binary code analysis. In particular, pre-trained embeddings of assembly code have become a gold standard for solving analysis tasks, such as measuring code similarity or recognizing functions. These embeddings are capable of learning a vector representation from unlabeled code. In contrast to natural language processing, however, label information is not scarce for many tasks in binary code analysis. For example, labeled training data for function boundaries, optimization levels, and argument types can be easily derived from debug information provided by a compiler. Consequently, the main motivation of embeddings does not transfer directly to binary code analysis. In this paper, we explore the role of pre-trained embeddings from a critical perspective. To this end, we systematically evaluate recent embeddings for assembly code on five downstream tasks using a corpus of 1.2 million functions from the Debian distribution. We observe that several embeddings perform similarly when sufficient labeled data is available, and that differences reported in prior work are hardly noticeable. Surprisingly, we find that end-to-end learning without pre-training performs best on average, which calls into question the need for specialized embeddings. By varying the amount of labeled data, we eventually derive guidelines for when embeddings offer advantages and when end-to-end learning is preferable for binary code analysis.
2502.08683
A Deep Learning approach for parametrized and time dependent Partial Differential Equations using Dimensionality Reduction and Neural ODEs
cs.LG
Partial Differential Equations (PDEs) are central to science and engineering. Since solving them is computationally expensive, a lot of effort has been put into approximating their solution operator via both traditional and recently increasingly Deep Learning (DL) techniques. A conclusive methodology capable of accounting both for (continuous) time and parameter dependency in such DL models however is still lacking. In this paper, we propose an autoregressive and data-driven method using the analogy with classical numerical solvers for time-dependent, parametric and (typically) nonlinear PDEs. We present how Dimensionality Reduction (DR) can be coupled with Neural Ordinary Differential Equations (NODEs) in order to learn the solution operator of arbitrary PDEs. The idea of our work is that it is possible to map the high-fidelity (i.e., high-dimensional) PDE solution space into a reduced (low-dimensional) space, which subsequently exhibits dynamics governed by a (latent) Ordinary Differential Equation (ODE). Solving this (easier) ODE in the reduced space allows avoiding solving the PDE in the high-dimensional solution space, thus decreasing the computational burden for repeated calculations for e.g., uncertainty quantification or design optimization purposes. The main outcome of this work is the importance of exploiting DR as opposed to the recent trend of building large and complex architectures: we show that by leveraging DR we can deliver not only more accurate predictions, but also a considerably lighter and faster DL model compared to existing methodologies.
2502.08684
Self-Evaluation for Job-Shop Scheduling
cs.LG cs.AI
Combinatorial optimization problems, such as scheduling and route planning, are crucial in various industries but are computationally intractable due to their NP-hard nature. Neural Combinatorial Optimization methods leverage machine learning to address these challenges but often depend on sequential decision-making, which is prone to error accumulation as small mistakes propagate throughout the process. Inspired by self-evaluation techniques in Large Language Models, we propose a novel framework that generates and evaluates subsets of assignments, moving beyond traditional stepwise approaches. Applied to the Job-Shop Scheduling Problem, our method integrates a heterogeneous graph neural network with a Transformer to build a policy model and a self-evaluation function. Experimental validation on challenging, well-known benchmarks demonstrates the effectiveness of our approach, surpassing state-of-the-art methods.
2502.08685
Beyond Models! Explainable Data Valuation and Metric Adaption for Recommendation
cs.LG cs.AI
User behavior records serve as the foundation for recommender systems. While the behavior data exhibits ease of acquisition, it often suffers from varying quality. Current methods employ data valuation to discern high-quality data from low-quality data. However, they tend to employ black-box design, lacking transparency and interpretability. Besides, they are typically tailored to specific evaluation metrics, leading to limited generality across various tasks. To overcome these issues, we propose an explainable and versatile framework DVR which can enhance the efficiency of data utilization tailored to any requirements of the model architectures and evaluation metrics. For explainable data valuation, a data valuator is presented to evaluate the data quality via calculating its Shapley value from the game-theoretic perspective, ensuring robust mathematical properties and reliability. In order to accommodate various evaluation metrics, including differentiable and non-differentiable ones, a metric adapter is devised based on reinforcement learning, where a metric is treated as the reinforcement reward that guides model optimization. Extensive experiments conducted on various benchmarks verify that our framework can improve the performance of current recommendation algorithms on various metrics including ranking accuracy, diversity, and fairness. Specifically, our framework achieves up to 34.7\% improvements over existing methods in terms of representative NDCG metric. The code is available at https://github.com/renqii/DVR.
2502.08686
EEG Artifact Detection and Correction with Deep Autoencoders
cs.LG cs.AI
EEG signals convey important information about brain activity both in healthy and pathological conditions. However, they are inherently noisy, which poses significant challenges for accurate analysis and interpretation. Traditional EEG artifact removal methods, while effective, often require extensive expert intervention. This study presents LSTEEG, a novel LSTM-based autoencoder designed for the detection and correction of artifacts in EEG signals. Leveraging deep learning, particularly LSTM layers, LSTEEG captures non-linear dependencies in sequential EEG data. LSTEEG demonstrates superior performance in both artifact detection and correction tasks compared to other state-of-the-art convolutional autoencoders. Our methodology enhances the interpretability and utility of the autoencoder's latent space, enabling data-driven automated artefact removal in EEG its application in downstream tasks. This research advances the field of efficient and accurate multi-channel EEG preprocessing, and promotes the implementation and usage of automated EEG analysis pipelines for brain health applications.
2502.08687
Data Augmentation to Improve Large Language Models in Food Hazard and Product Detection
cs.CL
The primary objective of this study is to demonstrate the impact of data augmentation using ChatGPT-4o-mini on food hazard and product analysis. The augmented data is generated using ChatGPT-4o-mini and subsequently used to train two large language models: RoBERTa-base and Flan-T5-base. The models are evaluated on test sets. The results indicate that using augmented data helped improve model performance across key metrics, including recall, F1 score, precision, and accuracy, compared to using only the provided dataset. The full code, including model training and the augmented dataset, can be found in this repository: https://github.com/AREEG94FAHAD/food-hazard-prdouct-cls
2502.08688
FAST: A Future Aircraft Sizing Tool for Advanced Aircraft and Propulsion System Design
cs.CE physics.comp-ph
Without radical technological advancements, the global aviation industry will continue to be a major carbon emitter. To reduce aviation's carbon emissions, innovative aircraft technology, including electrified aircraft propulsion, is under development. However, current aircraft sizing tools require detailed design information that may not be available early in the development process, particularly for novel technologies. This can yield suboptimal designs and inhibits innovation. A computational tool is needed to easily and rapidly size an aircraft configuration while allowing the designer to explore the design space, examine tradeoffs, and evaluate alternative designs. The Future Aircraft Sizing Tool (FAST), developed in Matlab, addresses this challenge by rapidly sizing aircraft with any propulsion architecture, including conventional, electric, and hybrid electric systems, even with limited initial data. FAST enables engineers to explore various aircraft configurations, evaluate design alternatives, assess performance across a flight envelope, and visualize concepts during the sizing process. By supporting early stage design, FAST addresses a gap in currently available computational tools for developing sustainable aviation technologies to help reduce the industry's carbon footprint.
2502.08689
Advancing machine fault diagnosis: A detailed examination of convolutional neural networks
cs.LG cs.AI
The growing complexity of machinery and the increasing demand for operational efficiency and safety have driven the development of advanced fault diagnosis techniques. Among these, convolutional neural networks (CNNs) have emerged as a powerful tool, offering robust and accurate fault detection and classification capabilities. This comprehensive review delves into the application of CNNs in machine fault diagnosis, covering its theoretical foundation, architectural variations, and practical implementations. The strengths and limitations of CNNs are analyzed in this domain, discussing their effectiveness in handling various fault types, data complexities, and operational environments. Furthermore, we explore the evolving landscape of CNN-based fault diagnosis, examining recent advancements in data augmentation, transfer learning, and hybrid architectures. Finally, we highlight future research directions and potential challenges to further enhance the application of CNNs for reliable and proactive machine fault diagnosis.
2502.08690
Skrr: Skip and Re-use Text Encoder Layers for Memory Efficient Text-to-Image Generation
cs.LG cs.AI cs.CV
Large-scale text encoders in text-to-image (T2I) diffusion models have demonstrated exceptional performance in generating high-quality images from textual prompts. Unlike denoising modules that rely on multiple iterative steps, text encoders require only a single forward pass to produce text embeddings. However, despite their minimal contribution to total inference time and floating-point operations (FLOPs), text encoders demand significantly higher memory usage, up to eight times more than denoising modules. To address this inefficiency, we propose Skip and Re-use layers (Skrr), a simple yet effective pruning strategy specifically designed for text encoders in T2I diffusion models. Skrr exploits the inherent redundancy in transformer blocks by selectively skipping or reusing certain layers in a manner tailored for T2I tasks, thereby reducing memory consumption without compromising performance. Extensive experiments demonstrate that Skrr maintains image quality comparable to the original model even under high sparsity levels, outperforming existing blockwise pruning methods. Furthermore, Skrr achieves state-of-the-art memory efficiency while preserving performance across multiple evaluation metrics, including the FID, CLIP, DreamSim, and GenEval scores.
2502.08691
AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Society
cs.SI cs.AI
Understanding human behavior and society is a central focus in social sciences, with the rise of generative social science marking a significant paradigmatic shift. By leveraging bottom-up simulations, it replaces costly and logistically challenging traditional experiments with scalable, replicable, and systematic computational approaches for studying complex social dynamics. Recent advances in large language models (LLMs) have further transformed this research paradigm, enabling the creation of human-like generative social agents and realistic simulacra of society. In this paper, we propose AgentSociety, a large-scale social simulator that integrates LLM-driven agents, a realistic societal environment, and a powerful large-scale simulation engine. Based on the proposed simulator, we generate social lives for over 10k agents, simulating their 5 million interactions both among agents and between agents and their environment. Furthermore, we explore the potential of AgentSociety as a testbed for computational social experiments, focusing on four key social issues: polarization, the spread of inflammatory messages, the effects of universal basic income policies, and the impact of external shocks such as hurricanes. These four issues serve as valuable cases for assessing AgentSociety's support for typical research methods -- such as surveys, interviews, and interventions -- as well as for investigating the patterns, causes, and underlying mechanisms of social issues. The alignment between AgentSociety's outcomes and real-world experimental results not only demonstrates its ability to capture human behaviors and their underlying mechanisms, but also underscores its potential as an important platform for social scientists and policymakers.
2502.08692
Efficient Split Learning LSTM Models for FPGA-based Edge IoT Devices
cs.LG cs.DC
Split Learning (SL) recently emerged as an efficient paradigm for distributed Machine Learning (ML) suitable for the Internet Of Things (IoT)-Cloud systems. However, deploying SL on resource-constrained edge IoT platforms poses a significant challenge in terms of balancing the model performance against the processing, memory, and energy resources. In this work, we present a practical study of deploying SL framework on a real-world Field-Programmable Gate Array (FPGA)-based edge IoT platform. We address the SL framework applied to a time-series processing model based on Recurrent Neural Networks (RNNs). Set in the context of river water quality monitoring and using real-world data, we train, optimize, and deploy a Long Short-Term Memory (LSTM) model on a given edge IoT FPGA platform in different SL configurations. Our results demonstrate the importance of aligning design choices with specific application requirements, whether it is maximizing speed, minimizing power, or optimizing for resource constraints.
2502.08695
A Bayesian Nonparametric Perspective on Mahalanobis Distance for Out of Distribution Detection
stat.ML cs.LG
Bayesian nonparametric methods are naturally suited to the problem of out-of-distribution (OOD) detection. However, these techniques have largely been eschewed in favor of simpler methods based on distances between pre-trained or learned embeddings of data points. Here we show a formal relationship between Bayesian nonparametric models and the relative Mahalanobis distance score (RMDS), a commonly used method for OOD detection. Building on this connection, we propose Bayesian nonparametric mixture models with hierarchical priors that generalize the RMDS. We evaluate these models on the OpenOOD detection benchmark and show that Bayesian nonparametric methods can improve upon existing OOD methods, especially in regimes where training classes differ in their covariance structure and where there are relatively few data points per class.
2502.08696
Scalable Discrete Diffusion Samplers: Combinatorial Optimization and Statistical Physics
cs.LG cond-mat.stat-mech cs.AI physics.comp-ph stat.ML
Learning to sample from complex unnormalized distributions over discrete domains emerged as a promising research direction with applications in statistical physics, variational inference, and combinatorial optimization. Recent work has demonstrated the potential of diffusion models in this domain. However, existing methods face limitations in memory scaling and thus the number of attainable diffusion steps since they require backpropagation through the entire generative process. To overcome these limitations we introduce two novel training methods for discrete diffusion samplers, one grounded in the policy gradient theorem and the other one leveraging Self-Normalized Neural Importance Sampling (SN-NIS). These methods yield memory-efficient training and achieve state-of-the-art results in unsupervised combinatorial optimization. Numerous scientific applications additionally require the ability of unbiased sampling. We introduce adaptations of SN-NIS and Neural Markov Chain Monte Carlo that enable for the first time the application of discrete diffusion models to this problem. We validate our methods on Ising model benchmarks and find that they outperform popular autoregressive approaches. Our work opens new avenues for applying diffusion models to a wide range of scientific applications in discrete domains that were hitherto restricted to exact likelihood models.
2502.08697
Bilevel Learning for Bilevel Planning
cs.RO
A robot that learns from demonstrations should not just imitate what it sees -- it should understand the high-level concepts that are being demonstrated and generalize them to new tasks. Bilevel planning is a hierarchical model-based approach where predicates (relational state abstractions) can be leveraged to achieve compositional generalization. However, previous bilevel planning approaches depend on predicates that are either hand-engineered or restricted to very simple forms, limiting their scalability to sophisticated, high-dimensional state spaces. To address this limitation, we present IVNTR, the first bilevel planning approach capable of learning neural predicates directly from demonstrations. Our key innovation is a neuro-symbolic bilevel learning framework that mirrors the structure of bilevel planning. In IVNTR, symbolic learning of the predicate "effects" and neural learning of the predicate "functions" alternate, with each providing guidance for the other. We evaluate IVNTR in six diverse robot planning domains, demonstrating its effectiveness in abstracting various continuous and high-dimensional states. While most existing approaches struggle to generalize (with <35% success rate), our IVNTR achieves an average of 77% success rate on unseen tasks. Additionally, we showcase IVNTR on a mobile manipulator, where it learns to perform real-world mobile manipulation tasks and generalizes to unseen test scenarios that feature new objects, new states, and longer task horizons. Our findings underscore the promise of learning and planning with abstractions as a path towards high-level generalization.
2502.08728
A Comparative Study of Machine Learning Algorithms for Stock Price Prediction Using Insider Trading Data
cs.LG
The research paper empirically investigates several machine learning algorithms to forecast stock prices depending on insider trading information. Insider trading offers special insights into market sentiment, pointing to upcoming changes in stock prices. This study examines the effectiveness of algorithms like decision trees, random forests, support vector machines (SVM) with different kernels, and K-Means Clustering using a dataset of Tesla stock transactions. Examining past data from April 2020 to March 2023, this study focuses on how well these algorithms identify trends and forecast stock price fluctuations. The paper uses Recursive Feature Elimination (RFE) and feature importance analysis to optimize the feature set and, hence, increase prediction accuracy. While it requires substantially greater processing time than other models, SVM with the Radial Basis Function (RBF) kernel displays the best accuracy. This paper highlights the trade-offs between accuracy and efficiency in machine learning models and proposes the possibility of pooling multiple data sources to raise prediction performance. The results of this paper aim to help financial analysts and investors in choosing strong algorithms to optimize investment strategies.
2502.08729
Policy Selection and Schedules for Exclusive Bus Lane and High Occupancy Vehicle Lane in a Bi-modal Transportation Corridor
math.OC cs.SY eess.SY
Efficient management of transportation corridors is critical for sustaining urban mobility, directly influencing transportation efficiency. Two prominent strategies for enhancing public transit services and alleviating congestion, Exclusive Bus Lane (EBL) and High Occupancy Vehicle Lane (HOVL), are gaining increasing attention. EBLs prioritize bus transit by providing dedicated lanes for faster travel times, while HOVLs encourage carpooling by reserving lanes for high-occupancy vehicles. However, static implementations of these policies may underutilize road resources and disrupt general-purpose lanes. Dynamic control of these policies, based on real-time demand, can potentially maximize road efficiency and minimize negative impacts. This study develops cost functions for Mixed Traffic Policy (MTP), Exclusive Bus Lane Policy (EBLP), and High Occupancy Vehicle Lane Policy (HOVLP), incorporating optimized bus frequency and demand split under equilibrium condition. Switching thresholds for policy selection are derived to identify optimal periods for implementing each policy based on dynamic demand simulated using an Ornstein-Uhlenbeck (O-U) process. Results reveal significant reductions in total system costs with the proposed dynamic policy integration. Compared to static implementations, the combined policy achieves cost reductions of 12.0%, 5.3% and 42.5% relative to MTP-only, EBLP-only, and HOVLP-only scenarios, respectively. Additionally, in two real case studies of existing EBL and HOVL operations, the proposed dynamic policy reduces total costs by 32.2% and 27.9%, respectively. The findings provide valuable insights for policymakers and transit planners, offering a robust framework for dynamically scheduling and integrating EBL and HOVL policies to optimize urban corridor efficiency and reduce overall system costs.
2502.08730
New Bounds for Sparse Variational Gaussian Processes
cs.LG stat.ME stat.ML
Sparse variational Gaussian processes (GPs) construct tractable posterior approximations to GP models. At the core of these methods is the assumption that the true posterior distribution over training function values ${\bf f}$ and inducing variables ${\bf u}$ is approximated by a variational distribution that incorporates the conditional GP prior $p({\bf f} | {\bf u})$ in its factorization. While this assumption is considered as fundamental, we show that for model training we can relax it through the use of a more general variational distribution $q({\bf f} | {\bf u})$ that depends on $N$ extra parameters, where $N$ is the number of training examples. In GP regression, we can analytically optimize the evidence lower bound over the extra parameters and express a tractable collapsed bound that is tighter than the previous bound. The new bound is also amenable to stochastic optimization and its implementation requires minor modifications to existing sparse GP code. Further, we also describe extensions to non-Gaussian likelihoods. On several datasets we demonstrate that our method can reduce bias when learning the hyperpaparameters and can lead to better predictive performance.
2502.08731
Equity-aware Design and Timing of Fare-free Transit Zoning under Demand Uncertainty
math.OC cs.SY eess.SY
We propose the first analytical stochastic model for optimizing the configuration and implementation policies of fare-free transit. The model focuses on a transportation corridor with two transportation modes: automobiles and buses. The corridor is divided into two sections, an inner one with fare-free transit service and an outer one with fare-based transit service. Under the static version of the model, the optimized length and frequency of the fare-free transit zone can be determined by maximizing total social welfare. The findings indicate that implementing fare-free transit can increase transit ridership and reduce automobile use within the fare-free zone while social equity among the demand groups can be enhanced by lengthening the fare-free zone. Notably, the optimal zone length increases when both social welfare and equity are considered jointly, compared to only prioritizing social welfare. The dynamic model, framed within a market entry and exit real options approach, solves the fare policy switching problem, establishing optimal timing policies for activating or terminating fare-free service. The results from dynamic models reveal earlier implementation and extended durations of fare-free transit in the social welfare-aware regime, driven by lower thresholds compared to the social equity-aware regime.
2502.08736
Recurrent Memory for Online Interdomain Gaussian Processes
cs.LG stat.ML
We propose a novel online Gaussian process (GP) model that is capable of capturing long-term memory in sequential data in an online regression setting. Our model, Online HiPPO Sparse Variational Gaussian Process Regression (OHSGPR), leverages the HiPPO (High-order Polynomial Projection Operators) framework, which is popularized in the RNN domain due to its long-range memory modeling capabilities. We interpret the HiPPO time-varying orthogonal projections as inducing variables with time-dependent orthogonal polynomial basis functions, which allows the SGPR inducing points to memorize the process history. We show that the HiPPO framework fits naturally into the interdomain GP framework and demonstrate that the kernel matrices can also be updated online in a recurrence form based on the ODE evolution of HiPPO. We evaluate our method on time series regression tasks, showing that it outperforms the existing online GP method in terms of predictive performance and computational efficiency
2502.08742
"Active Neighbour": A Novel Monitoring Model for Cyber-Physical Systems
eess.SY cs.SY
Over the past decade, advancements in technology have enabled Cyber-Physical Systems (CPS) to monitor sensor networks through various methodologies. However, these developments have concurrently introduced significant security challenges, necessitating robust protective measures. As a result, securing CPS has become a critical area of research. This paper reviews existing CPS monitoring models and introduces an innovative role-based monitoring model designed to meet contemporary security requirements. The proposed model is implemented within the COOJA simulator of the Contiki OS and evaluated under three distinct security configurations. Preliminary results demonstrate promising outcomes, although further comprehensive testing is ongoing.
2502.08744
Are Expressions for Music Emotions the Same Across Cultures?
cs.CL cs.HC cs.SD eess.AS
Music evokes profound emotions, yet the universality of emotional descriptors across languages remains debated. A key challenge in cross-cultural research on music emotion is biased stimulus selection and manual curation of taxonomies, predominantly relying on Western music and languages. To address this, we propose a balanced experimental design with nine online experiments in Brazil, the US, and South Korea, involving N=672 participants. First, we sample a balanced set of popular music from these countries. Using an open-ended tagging pipeline, we then gather emotion terms to create culture-specific taxonomies. Finally, using these bottom-up taxonomies, participants rate emotions of each song. This allows us to map emotional similarities within and across cultures. Results show consistency in high arousal, high valence emotions but greater variability in others. Notably, machine translations were often inadequate to capture music-specific meanings. These findings together highlight the need for a domain-sensitive, open-ended, bottom-up emotion elicitation approach to reduce cultural biases in emotion research.
2502.08745
IHEval: Evaluating Language Models on Following the Instruction Hierarchy
cs.CL
The instruction hierarchy, which establishes a priority order from system messages to user messages, conversation history, and tool outputs, is essential for ensuring consistent and safe behavior in language models (LMs). Despite its importance, this topic receives limited attention, and there is a lack of comprehensive benchmarks for evaluating models' ability to follow the instruction hierarchy. We bridge this gap by introducing IHEval, a novel benchmark comprising 3,538 examples across nine tasks, covering cases where instructions in different priorities either align or conflict. Our evaluation of popular LMs highlights their struggle to recognize instruction priorities. All evaluated models experience a sharp performance decline when facing conflicting instructions, compared to their original instruction-following performance. Moreover, the most competitive open-source model only achieves 48% accuracy in resolving such conflicts. Our results underscore the need for targeted optimization in the future development of LMs.
2502.08754
HistoSmith: Single-Stage Histology Image-Label Generation via Conditional Latent Diffusion for Enhanced Cell Segmentation and Classification
cs.CV cs.AI
Precise segmentation and classification of cell instances are vital for analyzing the tissue microenvironment in histology images, supporting medical diagnosis, prognosis, treatment planning, and studies of brain cytoarchitecture. However, the creation of high-quality annotated datasets for training remains a major challenge. This study introduces a novel single-stage approach (HistoSmith) for generating image-label pairs to augment histology datasets. Unlike state-of-the-art methods that utilize diffusion models with separate components for label and image generation, our approach employs a latent diffusion model to learn the joint distribution of cellular layouts, classification masks, and histology images. This model enables tailored data generation by conditioning on user-defined parameters such as cell types, quantities, and tissue types. Trained on the Conic H&E histopathology dataset and the Nissl-stained CytoDArk0 dataset, the model generates realistic and diverse labeled samples. Experimental results demonstrate improvements in cell instance segmentation and classification, particularly for underrepresented cell types like neutrophils in the Conic dataset. These findings underscore the potential of our approach to address data scarcity challenges.
2502.08756
From PowerPoint UI Sketches to Web-Based Applications: Pattern-Driven Code Generation for GIS Dashboard Development Using Knowledge-Augmented LLMs, Context-Aware Visual Prompting, and the React Framework
cs.AI cs.SE
Developing web-based GIS applications, commonly known as CyberGIS dashboards, for querying and visualizing GIS data in environmental research often demands repetitive and resource-intensive efforts. While Generative AI offers automation potential for code generation, it struggles with complex scientific applications due to challenges in integrating domain knowledge, software engineering principles, and UI design best practices. This paper introduces a knowledge-augmented code generation framework that retrieves software engineering best practices, domain expertise, and advanced technology stacks from a specialized knowledge base to enhance Generative Pre-trained Transformers (GPT) for front-end development. The framework automates the creation of GIS-based web applications (e.g., dashboards, interfaces) from user-defined UI wireframes sketched in tools like PowerPoint or Adobe Illustrator. A novel Context-Aware Visual Prompting method, implemented in Python, extracts layouts and interface features from these wireframes to guide code generation. Our approach leverages Large Language Models (LLMs) to generate front-end code by integrating structured reasoning, software engineering principles, and domain knowledge, drawing inspiration from Chain-of-Thought (CoT) prompting and Retrieval-Augmented Generation (RAG). A case study demonstrates the framework's capability to generate a modular, maintainable web platform hosting multiple dashboards for visualizing environmental and energy data (e.g., time-series, shapefiles, rasters) from user-sketched wireframes. By employing a knowledge-driven approach, the framework produces scalable, industry-standard front-end code using design patterns such as Model-View-ViewModel (MVVM) and frameworks like React. This significantly reduces manual effort in design and coding, pioneering an automated and efficient method for developing smart city software.
2502.08757
A Low-Complexity Plug-and-Play Deep Learning Model for Massive MIMO Precoding Across Sites
eess.SP cs.LG
Massive multiple-input multiple-output (mMIMO) technology has transformed wireless communication by enhancing spectral efficiency and network capacity. This paper proposes a novel deep learning-based mMIMO precoder to tackle the complexity challenges of existing approaches, such as weighted minimum mean square error (WMMSE), while leveraging meta-learning domain generalization and a teacher-student architecture to improve generalization across diverse communication environments. When deployed to a previously unseen site, the proposed model achieves excellent sum-rate performance while maintaining low computational complexity by avoiding matrix inversions and by using a simpler neural network structure. The model is trained and tested on a custom ray-tracing dataset composed of several base station locations. The experimental results indicate that our method effectively balances computational efficiency with high sum-rate performance while showcasing strong generalization performance in unseen environments. Furthermore, with fine-tuning, the proposed model outperforms WMMSE across all tested sites and SNR conditions while reducing complexity by at least 73$\times$.
2502.08758
Compression of Site-Specific Deep Neural Networks for Massive MIMO Precoding
eess.SP cs.LG
The deployment of deep learning (DL) models for precoding in massive multiple-input multiple-output (mMIMO) systems is often constrained by high computational demands and energy consumption. In this paper, we investigate the compute energy efficiency of mMIMO precoders using DL-based approaches, comparing them to conventional methods such as zero forcing and weighted minimum mean square error (WMMSE). Our energy consumption model accounts for both memory access and calculation energy within DL accelerators. We propose a framework that incorporates mixed-precision quantization-aware training and neural architecture search to reduce energy usage without compromising accuracy. Using a ray-tracing dataset covering various base station sites, we analyze how site-specific conditions affect the energy efficiency of compressed models. Our results show that deep neural network compression generates precoders with up to 35 times higher energy efficiency than WMMSE at equal performance, depending on the scenario and the desired rate. These results establish a foundation and a benchmark for the development of energy-efficient DL-based mMIMO precoders.
2502.08759
Contextual bandits with entropy-based human feedback
cs.AI
In recent years, preference-based human feedback mechanisms have become essential for enhancing model performance across diverse applications, including conversational AI systems such as ChatGPT. However, existing approaches often neglect critical aspects, such as model uncertainty and the variability in feedback quality. To address these challenges, we introduce an entropy-based human feedback framework for contextual bandits, which dynamically balances exploration and exploitation by soliciting expert feedback only when model entropy exceeds a predefined threshold. Our method is model-agnostic and can be seamlessly integrated with any contextual bandit agent employing stochastic policies. Through comprehensive experiments, we show that our approach achieves significant performance improvements while requiring minimal human feedback, even under conditions of suboptimal feedback quality. This work not only presents a novel strategy for feedback solicitation but also highlights the robustness and efficacy of incorporating human guidance into machine learning systems. Our code is publicly available: https://github.com/BorealisAI/CBHF
2502.08764
Demand Response Optimization MILP Framework for Microgrids with DERs
eess.SY cs.LG cs.SY
The integration of renewable energy sources in microgrids introduces significant operational challenges due to their intermittent nature and the mismatch between generation and demand patterns. Effective demand response (DR) strategies are crucial for maintaining system stability and economic efficiency, particularly in microgrids with high renewable penetration. This paper presents a comprehensive mixed-integer linear programming (MILP) framework for optimizing DR operations in a microgrid with solar generation and battery storage systems. The framework incorporates load classification, dynamic price thresholding, and multi-period coordination for optimal DR event scheduling. Analysis across seven distinct operational scenarios demonstrates consistent peak load reduction of 10\% while achieving energy cost savings ranging from 13.1\% to 38.0\%. The highest performance was observed in scenarios with high solar generation, where the framework achieved 38.0\% energy cost reduction through optimal coordination of renewable resources and DR actions. The results validate the framework's effectiveness in managing diverse operational challenges while maintaining system stability and economic efficiency.
2502.08766
Unlocking Mental Health: Exploring College Students' Well-being through Smartphone Behaviors
cs.CY cs.HC cs.LG cs.SE
The global mental health crisis is a pressing concern, with college students particularly vulnerable to rising mental health disorders. The widespread use of smartphones among young adults, while offering numerous benefits, has also been linked to negative outcomes such as addiction and regret, significantly impacting well-being. Leveraging the longest longitudinal dataset collected over four college years through passive mobile sensing, this study is the first to examine the relationship between students' smartphone unlocking behaviors and their mental health at scale in real-world settings. We provide the first evidence demonstrating the predictability of phone unlocking behaviors for mental health outcomes based on a large dataset, highlighting the potential of these novel features for future predictive models. Our findings reveal important variations in smartphone usage across genders and locations, offering a deeper understanding of the interplay between digital behaviors and mental health. We highlight future research directions aimed at mitigating adverse effects and promoting digital well-being in this population.
2502.08767
SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence
cs.CL cs.AI
Providing Language Models (LMs) with relevant evidence in the context (either via retrieval or user-provided) can significantly improve their ability to provide factually correct grounded responses. However, recent studies have found that LMs often struggle to fully comprehend and utilize key evidence from the context, especially when it contains noise and irrelevant information - an issue common in real-world scenarios. To address this, we propose SelfElicit, an inference-time approach that helps LMs focus on key contextual evidence through self-guided explicit highlighting. By leveraging the inherent evidence-finding capabilities of LMs using the attention scores of deeper layers, our method automatically identifies and emphasizes key evidence within the input context, facilitating more accurate and factually grounded responses without additional training or iterative prompting. We demonstrate that SelfElicit brings consistent and significant improvement on multiple evidence-based QA tasks for various LM families while maintaining computational efficiency. Our code and documentation are available at https://github.com/ZhiningLiu1998/SelfElicit.
2502.08769
Cluster and Predict Latent Patches for Improved Masked Image Modeling
cs.CV cs.AI
Masked Image Modeling (MIM) offers a promising approach to self-supervised representation learning, however existing MIM models still lag behind the state-of-the-art. In this paper, we systematically analyze target representations, loss functions, and architectures, to introduce CAPI - a novel pure-MIM framework that relies on the prediction of latent clusterings. Our approach leverages a clustering-based loss, which is stable to train, and exhibits promising scaling properties. Our ViT-L backbone, CAPI, achieves 83.8% accuracy on ImageNet and 32.1% mIoU on ADE20K with simple linear probes, substantially outperforming previous MIM methods and approaching the performance of the current state-of-the-art, DINOv2. We release all our code and models.
2502.08773
Universal Model Routing for Efficient LLM Inference
cs.CL cs.LG
Large language models' significant advances in capabilities are accompanied by significant increases in inference costs. Model routing is a simple technique for reducing inference cost, wherein one maintains a pool of candidate LLMs, and learns to route each prompt to the smallest feasible LLM. Existing works focus on learning a router for a fixed pool of LLMs. In this paper, we consider the problem of dynamic routing, where new, previously unobserved LLMs are available at test time. We propose a new approach to this problem that relies on representing each LLM as a feature vector, derived based on predictions on a set of representative prompts. Based on this, we detail two effective strategies, relying on cluster-based routing and a learned cluster map respectively. We prove that these strategies are estimates of a theoretically optimal routing rule, and provide an excess risk bound to quantify their errors. Experiments on a range of public benchmarks show the effectiveness of the proposed strategies in routing amongst more than 30 unseen LLMs.
2502.08774
Exploring Test Time Adaptation for Subcortical Segmentation of the Fetal Brain in 3D Ultrasound
cs.CV cs.AI cs.LG
Monitoring the growth of subcortical regions of the fetal brain in ultrasound (US) images can help identify the presence of abnormal development. Manually segmenting these regions is a challenging task, but recent work has shown that it can be automated using deep learning. However, applying pretrained models to unseen freehand US volumes often leads to a degradation of performance due to the vast differences in acquisition and alignment. In this work, we first demonstrate that test time adaptation (TTA) can be used to improve model performance in the presence of both real and simulated domain shifts. We further propose a novel TTA method by incorporating a normative atlas as a prior for anatomy. In the presence of various types of domain shifts, we benchmark the performance of different TTA methods and demonstrate the improvements brought by our proposed approach, which may further facilitate automated monitoring of fetal brain development. Our code is available at https://github.com/joshuaomolegan/TTA-for-3D-Fetal-Subcortical-Segmentation.
2502.08776
Treatment response as a latent variable
stat.ME cs.LG stat.ML
Scientists often need to analyze the samples in a study that responded to treatment in order to refine their hypotheses and find potential causal drivers of response. Natural variation in outcomes makes teasing apart responders from non-responders a statistical inference problem. To handle latent responses, we introduce the causal two-groups (C2G) model, a causal extension of the classical two-groups model. The C2G model posits that treated samples may or may not experience an effect, according to some prior probability. We propose two empirical Bayes procedures for the causal two-groups model, one under semi-parametric conditions and another under fully nonparametric conditions. The semi-parametric model assumes additive treatment effects and is identifiable from observed data. The nonparametric model is unidentifiable, but we show it can still be used to test for response in each treated sample. We show empirically and theoretically that both methods for selecting responders control the false discovery rate at the target level with near-optimal power. We also propose two novel estimands of interest and provide a strategy for deriving estimand intervals in the unidentifiable nonparametric model. On a cancer immunotherapy dataset, the nonparametric C2G model recovers clinically-validated predictive biomarkers of both positive and negative outcomes. Code is available at https://github.com/tansey-lab/causal2groups.
2502.08777
Zero-Shot Belief: A Hard Problem for LLMs
cs.CL
We present two LLM-based approaches to zero-shot source-and-target belief prediction on FactBank: a unified system that identifies events, sources, and belief labels in a single pass, and a hybrid approach that uses a fine-tuned DeBERTa tagger for event detection. We show that multiple open-sourced, closed-source, and reasoning-based LLMs struggle with the task. Using the hybrid approach, we achieve new state-of-the-art results on FactBank and offer a detailed error analysis. Our approach is then tested on the Italian belief corpus ModaFact.
2502.08779
SB-Bench: Stereotype Bias Benchmark for Large Multimodal Models
cs.CV
Stereotype biases in Large Multimodal Models (LMMs) perpetuate harmful societal prejudices, undermining the fairness and equity of AI applications. As LMMs grow increasingly influential, addressing and mitigating inherent biases related to stereotypes, harmful generations, and ambiguous assumptions in real-world scenarios has become essential. However, existing datasets evaluating stereotype biases in LMMs often lack diversity and rely on synthetic images, leaving a gap in bias evaluation for real-world visual contexts. To address this, we introduce the Stereotype Bias Benchmark (SB-bench), the most comprehensive framework to date for assessing stereotype biases across nine diverse categories with non-synthetic images. SB-bench rigorously evaluates LMMs through carefully curated, visually grounded scenarios, challenging them to reason accurately about visual stereotypes. It offers a robust evaluation framework featuring real-world visual samples, image variations, and multiple-choice question formats. By introducing visually grounded queries that isolate visual biases from textual ones, SB-bench enables a precise and nuanced assessment of a model's reasoning capabilities across varying levels of difficulty. Through rigorous testing of state-of-the-art open-source and closed-source LMMs, SB-bench provides a systematic approach to assessing stereotype biases in LMMs across key social dimensions. This benchmark represents a significant step toward fostering fairness in AI systems and reducing harmful biases, laying the groundwork for more equitable and socially responsible LMMs. Our code and dataset are publicly available.
2502.08782
A comparative study of different TSO-DSO coordination in the reserve market
eess.SY cs.SY
The increasing penetration of Distributed Energy Resources (DERs) in the distribution system has led to the emergence of a new market actor - the aggregator. The aggregator serves as a facilitator, enabling flexibility asset owners to get access to different markets. In which, EVs aggregators are gaining more attention due to their expanding use and potential to provide services in various types of markets, particularly in the reserve market. Currently, TSO indirectly utilizes these resources under the management of the distribution system operators (DSO), which can negatively impact the distribution grid. Conversely, adjustments from DSOs can impact service provision to TSO due to the shortage of TSO usage information. These factors highlight the importance of evaluating the service provision from aggregators under different TSO-DSO coordination schemes. This paper focuses on the provision of flexibility from electric vehicles (EVs) aggregators for balancing service in the TSO-DSO hybrid-managed and compares it with the DSO-managed coordination schemes. The behavior of aggregators reacting to price fluctuations and TSO requests under different coordination schemes and simulation scenarios is thoroughly evaluated. Additionally, their impact on the grid is analyzed through the DSO's congestion management process and validated using data from a real part of the Dutch distribution network. Results find that the hybrid-managed coordination scheme gives more benefit to the aggregator than the DSO-managed scheme and the EVs aggregator will gain more profit in winter than summer due to more upward regulation service is needed.
2502.08783
Learning Discontinuous Galerkin Solutions to Elliptic Problems via Small Linear Convolutional Neural Networks
cs.LG cs.NA math.NA
In recent years, there has been an increasing interest in using deep learning and neural networks to tackle scientific problems, particularly in solving partial differential equations (PDEs). However, many neural network-based methods, such as physics-informed neural networks, depend on automatic differentiation and the sampling of collocation points, which can result in a lack of interpretability and lower accuracy compared to traditional numerical methods. To address this issue, we propose two approaches for learning discontinuous Galerkin solutions to PDEs using small linear convolutional neural networks. Our first approach is supervised and depends on labeled data, while our second approach is unsupervised and does not rely on any training data. In both cases, our methods use substantially fewer parameters than similar numerics-based neural networks while also demonstrating comparable accuracy to the true and DG solutions for elliptic problems.
2502.08784
Acoustic Wave Manipulation Through Sparse Robotic Actuation
cs.RO cs.AI
Recent advancements in robotics, control, and machine learning have facilitated progress in the challenging area of object manipulation. These advancements include, among others, the use of deep neural networks to represent dynamics that are partially observed by robot sensors, as well as effective control using sparse control signals. In this work, we explore a more general problem: the manipulation of acoustic waves, which are partially observed by a robot capable of influencing the waves through spatially sparse actuators. This problem holds great potential for the design of new artificial materials, ultrasonic cutting tools, energy harvesting, and other applications. We develop an efficient data-driven method for robot learning that is applicable to either focusing scattered acoustic energy in a designated region or suppressing it, depending on the desired task. The proposed method is better in terms of a solution quality and computational complexity as compared to a state-of-the-art learning based method for manipulation of dynamical systems governed by partial differential equations. Furthermore our proposed method is competitive with a classical semi-analytical method in acoustics research on the demonstrated tasks. We have made the project code publicly available, along with a web page featuring video demonstrations: https://gladisor.github.io/waves/.
2502.08785
Decision Tree Based Wrappers for Hearing Loss
cs.LG cs.SD
Audiology entities are using Machine Learning (ML) models to guide their screening towards people at risk. Feature Engineering (FE) focuses on optimizing data for ML models, with evolutionary methods being effective in feature selection and construction tasks. This work aims to benchmark an evolutionary FE wrapper, using models based on decision trees as proxies. The FEDORA framework is applied to a Hearing Loss (HL) dataset, being able to reduce data dimensionality and statistically maintain baseline performance. Compared to traditional methods, FEDORA demonstrates superior performance, with a maximum balanced accuracy of 76.2%, using 57 features. The framework also generated an individual that achieved 72.8% balanced accuracy using a single feature.
2502.08786
MRUCT: Mixed Reality Assistance for Acupuncture Guided by Ultrasonic Computed Tomography
cs.HC cs.CV cs.GR
Chinese acupuncture practitioners primarily depend on muscle memory and tactile feedback to insert needles and accurately target acupuncture points, as the current workflow lacks imaging modalities and visual aids. Consequently, new practitioners often learn through trial and error, requiring years of experience to become proficient and earn the trust of patients. Medical students face similar challenges in mastering this skill. To address these challenges, we developed an innovative system, MRUCT, that integrates ultrasonic computed tomography (UCT) with mixed reality (MR) technology to visualize acupuncture points in real-time. This system offers offline image registration and real-time guidance during needle insertion, enabling them to accurately position needles based on anatomical structures such as bones, muscles, and auto-generated reference points, with the potential for clinical implementation. In this paper, we outline the non-rigid registration methods used to reconstruct anatomical structures from UCT data, as well as the key design considerations of the MR system. We evaluated two different 3D user interface (3DUI) designs and compared the performance of our system to traditional workflows for both new practitioners and medical students. The results highlight the potential of MR to enhance therapeutic medical practices and demonstrate the effectiveness of the system we developed.
2502.08788
If Multi-Agent Debate is the Answer, What is the Question?
cs.CL cs.LG
Multi-agent debate (MAD) has emerged as a promising approach to enhance the factual accuracy and reasoning quality of large language models (LLMs) by engaging multiple agents in iterative discussions during inference. Despite its potential, we argue that current MAD research suffers from critical shortcomings in evaluation practices, including limited dataset overlap and inconsistent baselines, raising significant concerns about generalizability. Correspondingly, this paper presents a systematic evaluation of five representative MAD methods across nine benchmarks using four foundational models. Surprisingly, our findings reveal that MAD methods fail to reliably outperform simple single-agent baselines such as Chain-of-Thought and Self-Consistency, even when consuming additional inference-time computation. From our analysis, we found that model heterogeneity can significantly improve MAD frameworks. We propose Heter-MAD enabling a single LLM agent to access the output from heterogeneous foundation models, which boosts the performance of current MAD frameworks. Finally, we outline potential directions for advancing MAD, aiming to spark a broader conversation and inspire future work in this area.
2502.08789
Delay Analysis of 5G HARQ in the Presence of Decoding and Feedback Latencies
cs.IT cs.SY eess.SY math.IT
The growing demand for stringent quality of service (QoS) guarantees in 5G networks requires accurate characterisation of delay performance, often measured using Delay Violation Probability (DVP) for a given target delay. Widely used retransmission schemes like Automatic Repeat reQuest (ARQ) and Hybrid ARQ (HARQ) improve QoS through effective feedback, incremental redundancy (IR), and parallel retransmission processes. However, existing works to quantify the DVP under these retransmission schemes overlook practical aspects such as decoding complexity, feedback delays, and the resulting need for multiple parallel ARQ/HARQ processes that enable packet transmissions without waiting for previous feedback, thus exploiting valuable transmission opportunities. This work proposes a comprehensive multi-server delay model for ARQ/HARQ that incorporates these aspects. Using a finite blocklength error model, we derive closed-form expressions and algorithms for accurate DVP evaluation under realistic 5G configurations aligned with 3GPP standards. Our numerical evaluations demonstrate notable improvements in DVP accuracy over the state-of-the-art, highlight the impact of parameter tuning and resource allocation, and reveal how DVP affects system throughput.
2502.08791
ClipRover: Zero-shot Vision-Language Exploration and Target Discovery by Mobile Robots
cs.RO
Vision-language navigation (VLN) has emerged as a promising paradigm, enabling mobile robots to perform zero-shot inference and execute tasks without specific pre-programming. However, current systems often separate map exploration and path planning, with exploration relying on inefficient algorithms due to limited (partially observed) environmental information. In this paper, we present a novel navigation pipeline named ''ClipRover'' for simultaneous exploration and target discovery in unknown environments, leveraging the capabilities of a vision-language model named CLIP. Our approach requires only monocular vision and operates without any prior map or knowledge about the target. For comprehensive evaluations, we design the functional prototype of a UGV (unmanned ground vehicle) system named ''Rover Master'', a customized platform for general-purpose VLN tasks. We integrate and deploy the ClipRover pipeline on Rover Master to evaluate its throughput, obstacle avoidance capability, and trajectory performance across various real-world scenarios. Experimental results demonstrate that ClipRover consistently outperforms traditional map traversal algorithms and achieves performance comparable to path-planning methods that depend on prior map and target knowledge. Notably, ClipRover offers real-time active navigation without requiring pre-captured candidate images or pre-built node graphs, addressing key limitations of existing VLN pipelines.
2502.08792
Auction Design using Value Prediction with Hallucinations
cs.GT cs.AI
We investigate a Bayesian mechanism design problem where a seller seeks to maximize revenue by selling an indivisible good to one of n buyers, incorporating potentially unreliable predictions (signals) of buyers' private values derived from a machine learning model. We propose a framework where these signals are sometimes reflective of buyers' true valuations but other times are hallucinations, which are uncorrelated with the buyers' true valuations. Our main contribution is a characterization of the optimal auction under this framework. Our characterization establishes a near-decomposition of how to treat types above and below the signal. For the one buyer case, the seller's optimal strategy is to post one of three fairly intuitive prices depending on the signal, which we call the "ignore", "follow" and "cap" actions.
2502.08794
Spectral Journey: How Transformers Predict the Shortest Path
cs.LG
Decoder-only transformers lead to a step-change in capability of large language models. However, opinions are mixed as to whether they are really planning or reasoning. A path to making progress in this direction is to study the model's behavior in a setting with carefully controlled data. Then interpret the learned representations and reverse-engineer the computation performed internally. We study decoder-only transformer language models trained from scratch to predict shortest paths on simple, connected and undirected graphs. In this setting, the representations and the dynamics learned by the model are interpretable. We present three major results: (1) Two-layer decoder-only language models can learn to predict shortest paths on simple, connected graphs containing up to 10 nodes. (2) Models learn a graph embedding that is correlated with the spectral decomposition of the line graph. (3) Following the insights, we discover a novel approximate path-finding algorithm Spectral Line Navigator (SLN) that finds shortest path by greedily selecting nodes in the space of spectral embedding of the line graph.
2502.08795
Low-Resolution Neural Networks
cs.LG
The expanding scale of large neural network models introduces significant challenges, driving efforts to reduce memory usage and enhance computational efficiency. Such measures are crucial to ensure the practical implementation and effective application of these sophisticated models across a wide array of use cases. This study examines the impact of parameter bit precision on model performance compared to standard 32-bit models, with a focus on multiclass object classification in images. The models analyzed include those with fully connected layers, convolutional layers, and transformer blocks, with model weight resolution ranging from 1 bit to 4.08 bits. The findings indicate that models with lower parameter bit precision achieve results comparable to 32-bit models, showing promise for use in memory-constrained devices. While low-resolution models with a small number of parameters require more training epochs to achieve accuracy comparable to 32-bit models, those with a large number of parameters achieve similar performance within the same number of epochs. Additionally, data augmentation can destabilize training in low-resolution models, but including zero as a potential value in the weight parameters helps maintain stability and prevents performance degradation. Overall, 2.32-bit weights offer the optimal balance of memory reduction, performance, and efficiency. However, further research should explore other dataset types and more complex and larger models. These findings suggest a potential new era for optimized neural network models with reduced memory requirements and improved computational efficiency, though advancements in dedicated hardware are necessary to fully realize this potential.
2502.08796
A Systematic Review on the Evaluation of Large Language Models in Theory of Mind Tasks
cs.CL cs.CY cs.HC
In recent years, evaluating the Theory of Mind (ToM) capabilities of large language models (LLMs) has received significant attention within the research community. As the field rapidly evolves, navigating the diverse approaches and methodologies has become increasingly complex. This systematic review synthesizes current efforts to assess LLMs' ability to perform ToM tasks, an essential aspect of human cognition involving the attribution of mental states to oneself and others. Despite notable advancements, the proficiency of LLMs in ToM remains a contentious issue. By categorizing benchmarks and tasks through a taxonomy rooted in cognitive science, this review critically examines evaluation techniques, prompting strategies, and the inherent limitations of LLMs in replicating human-like mental state reasoning. A recurring theme in the literature reveals that while LLMs demonstrate emerging competence in ToM tasks, significant gaps persist in their emulation of human cognitive abilities.
2502.08803
Deep EEG Super-Resolution: Upsampling EEG Spatial Resolution with Generative Adversarial Networks
cs.LG
Electroencephalography (EEG) activity contains a wealth of information about what is happening within the human brain. Recording more of this data has the potential to unlock endless future applications. However, the cost of EEG hardware is increasingly expensive based upon the number of EEG channels being recorded simultaneously. We combat this problem in this paper by proposing a novel deep EEG super-resolution (SR) approach based on Generative Adversarial Networks (GANs). This approach can produce high spatial resolution EEG data from low resolution samples, by generating channel-wise upsampled data to effectively interpolate numerous missing channels, thus reducing the need for expensive EEG equipment. We tested the performance using an EEG dataset from a mental imagery task. Our proposed GAN model provided 10^4 fold and 10^2 fold reduction in mean-squared error (MSE) and mean-absolute error (MAE), respectively, over the baseline bicubic interpolation method. We further validate our method by training a classifier on the original classification task, which displayed minimal loss in accuracy while using the super-resolved data. The proposed SR EEG by GAN is a promising approach to improve the spatial resolution of low density EEG headsets.
2502.08806
CLOVER: A Test Case Generation Benchmark with Coverage, Long-Context, and Verification
cs.SE cs.AI cs.LG
Software testing is a critical aspect of software development, yet generating test cases remains a routine task for engineers. This paper presents a benchmark, CLOVER, to evaluate models' capabilities in generating and completing test cases under specific conditions. Spanning from simple assertion completions to writing test cases that cover specific code blocks across multiple files, these tasks are based on 12 python repositories, analyzing 845 problems with context lengths ranging from 4k to 128k tokens. Utilizing code testing frameworks, we propose a method to construct retrieval contexts using coverage information. While models exhibit comparable performance with short contexts, notable differences emerge with 16k contexts. Notably, models like GPT-4o and Claude 3.5 can effectively leverage relevant snippets; however, all models score below 35\% on the complex Task III, even with the oracle context provided, underscoring the benchmark's significance and the potential for model improvement. The benchmark is containerized for code execution across tasks, and we will release the code, data, and construction methodologies.
2502.08807
InTAR: Inter-Task Auto-Reconfigurable Accelerator Design for High Data Volume Variation in DNNs
cs.AR cs.LG
The rise of deep neural networks (DNNs) has driven a boom in AI services, which results in an increased demand for computing power and memory. In modern DNNs, the data sizes produced and consumed are highly varied across operations (high data volume variation, HDV). Because existing design paradigms use fixed execution patterns that lead to either low computational efficiency due to pipeline stalls or frequent off-chip memory accesses to manage large intermediate data, HDV applications are challenging to accelerate on FPGAs. To address these challenges, we introduce the Inter-Task Auto-Reconfigurable Accelerator (InTAR), a novel accelerator design for HDV applications on FPGAs. InTAR combines the high computational efficiency of sequential execution with the reduced off-chip memory overhead of dataflow execution. It switches execution patterns automatically with a static schedule determined before circuit design based on resource constraints and model parameters. Unlike previous reconfigurable accelerators, InTAR encodes reconfiguration schedules during circuit design, allowing model-specific optimizations that allocate only the necessary logic and interconnects. Thus, InTAR achieves a high clock frequency with fewer resources and low reconfiguration time. Furthermore, InTAR supports high-level tools such as HLS for fast design generation. We implement a set of multi-task kernels in various HDV DNNs using InTAR. Compared with dataflow and sequential accelerators, InTAR exhibits $1.8\times$ and $7.1 \times$ speedups correspondingly. We also implement InTAR for GPT-2 medium as a more complex example, which achieves a speedup of $\mathbf{3.65 \sim 39.14\times}$ and a $\mathbf{1.72 \sim 10.44\times}$ boost in DSP efficiency compared to the corresponding SoTA accelerators on FPGAs.
2502.08808
A First-order Generative Bilevel Optimization Framework for Diffusion Models
cs.LG math.OC stat.ML
Diffusion models, which iteratively denoise data samples to synthesize high-quality outputs, have achieved empirical success across domains. However, optimizing these models for downstream tasks often involves nested bilevel structures, such as tuning hyperparameters for fine-tuning tasks or noise schedules in training dynamics, where traditional bilevel methods fail due to the infinite-dimensional probability space and prohibitive sampling costs. We formalize this challenge as a generative bilevel optimization problem and address two key scenarios: (1) fine-tuning pre-trained models via an inference-only lower-level solver paired with a sample-efficient gradient estimator for the upper level, and (2) training diffusion models from scratch with noise schedule optimization by reparameterizing the lower-level problem and designing a computationally tractable gradient estimator. Our first-order bilevel framework overcomes the incompatibility of conventional bilevel methods with diffusion processes, offering theoretical grounding and computational practicality. Experiments demonstrate that our method outperforms existing fine-tuning and hyperparameter search baselines.
2502.08813
Measuring Anxiety Levels with Head Motion Patterns in Severe Depression Population
cs.CV
Depression and anxiety are prevalent mental health disorders that frequently cooccur, with anxiety significantly influencing both the manifestation and treatment of depression. An accurate assessment of anxiety levels in individuals with depression is crucial to develop effective and personalized treatment plans. This study proposes a new noninvasive method for quantifying anxiety severity by analyzing head movements -- specifically speed, acceleration, and angular displacement -- during video-recorded interviews with patients suffering from severe depression. Using data from a new CALYPSO Depression Dataset, we extracted head motion characteristics and applied regression analysis to predict clinically evaluated anxiety levels. Our results demonstrate a high level of precision, achieving a mean absolute error (MAE) of 0.35 in predicting the severity of psychological anxiety based on head movement patterns. This indicates that our approach can enhance the understanding of anxiety's role in depression and assist psychiatrists in refining treatment strategies for individuals.
2502.08818
Lexical Manifold Reconfiguration in Large Language Models: A Novel Architectural Approach for Contextual Modulation
cs.CL
Contextual adaptation in token embeddings plays a central role in determining how well language models maintain coherence and retain semantic relationships over extended text sequences. Static embeddings often impose constraints on lexical flexibility, leading to suboptimal performance when faced with complex sentence structures or domain-specific terminology shifts. To address this limitation, a structured approach was developed for dynamically reconfiguring token embeddings through continuous geometric transformations, ensuring that representations evolved in response to evolving discourse structures. A manifold-based transformation mechanism was integrated to regulate lexical positioning, allowing embeddings to undergo controlled shifts while preserving linguistic relationships across varying textual contexts. Empirical evaluations demonstrated that embedding reconfiguration contributed to reductions in perplexity, improved lexical coherence, and enhanced sentence-level continuity, particularly in structured and domain-adaptive text generation tasks. Comparative analyses of embedding drift indicated that dynamically restructured representations maintained stronger contextual consistency, reducing misalignment in token dependencies while preserving fluency in language modeling outputs. Computational overhead assessments confirmed that while training complexity increased due to the iterative refinement of embeddings, inference remained efficient, ensuring practical feasibility for real-time generation. Evaluations across multiple datasets further demonstrated that dynamically modulated embeddings exhibited broader lexical diversity, reducing repetitive token patterns and enabling a more adaptable representation learning process.
2502.08820
Can a Single Model Master Both Multi-turn Conversations and Tool Use? CoALM: A Unified Conversational Agentic Language Model
cs.AI cs.CL
Large Language Models (LLMs) with API-calling capabilities enabled building effective Language Agents (LA), while also revolutionizing the conventional task-oriented dialogue (TOD) paradigm. However, current approaches face a critical dilemma: TOD systems are often trained on a limited set of target APIs, requiring new data to maintain their quality when interfacing with new services, while LAs are not trained to maintain user intent over multi-turn conversations. Because both robust multi-turn management and advanced function calling are crucial for effective conversational agents, we evaluate these skills on three popular benchmarks: MultiWOZ 2.4 (TOD), BFCL V3 (LA), and API-Bank (LA), and our analyses reveal that specialized approaches excel in one domain but underperform in the other. To bridge this chasm, we introduce CoALM (Conversational Agentic Language Model), a unified approach that integrates both conversational and agentic capabilities. We created CoALM-IT, a carefully constructed multi-task dataset that interleave multi-turn ReAct reasoning with complex API usage. Using CoALM-IT, we train three models CoALM 8B, CoALM 70B, and CoALM 405B, which outperform top domain-specific models, including GPT-4o, across all three benchmarks. This demonstrates the feasibility of a single model approach for both TOD and LA, setting a new standard for conversational agents.
2502.08821
DejAIvu: Identifying and Explaining AI Art on the Web in Real-Time with Saliency Maps
cs.CV cs.AI cs.LG
The recent surge in advanced generative models, such as diffusion models and generative adversarial networks (GANs), has led to an alarming rise in AI-generated images across various domains on the web. While such technologies offer benefits such as democratizing artistic creation, they also pose challenges in misinformation, digital forgery, and authenticity verification. Additionally, the uncredited use of AI-generated images in media and marketing has sparked significant backlash from online communities. In response to this, we introduce DejAIvu, a Chrome Web extension that combines real-time AI-generated image detection with saliency-based explainability while users browse the web. Using an ONNX-optimized deep learning model, DejAIvu automatically analyzes images on websites such as Google Images, identifies AI-generated content using model inference, and overlays a saliency heatmap to highlight AI-related artifacts. Our approach integrates efficient in-browser inference, gradient-based saliency analysis, and a seamless user experience, ensuring that AI detection is both transparent and interpretable. We also evaluate DejAIvu across multiple pretrained architectures and benchmark datasets, demonstrating high accuracy and low latency, making it a practical and deployable tool for enhancing AI image accountability. The code for this system can be found at https://github.com/Noodulz/dejAIvu.
2502.08822
$\mathsf{CSMAE~}$:~Cataract Surgical Masked Autoencoder (MAE) based Pre-training
cs.CV
Automated analysis of surgical videos is crucial for improving surgical training, workflow optimization, and postoperative assessment. We introduce a CSMAE, Masked Autoencoder (MAE)-based pretraining approach, specifically developed for Cataract Surgery video analysis, where instead of randomly selecting tokens for masking, they are selected based on the spatiotemporal importance of the token. We created a large dataset of cataract surgery videos to improve the model's learning efficiency and expand its robustness in low-data regimes. Our pre-trained model can be easily adapted for specific downstream tasks via fine-tuning, serving as a robust backbone for further analysis. Through rigorous testing on a downstream step-recognition task on two Cataract Surgery video datasets, D99 and Cataract-101, our approach surpasses current state-of-the-art self-supervised pretraining and adapter-based transfer learning methods by a significant margin. This advancement not only demonstrates the potential of our MAE-based pretraining in the field of surgical video analysis but also sets a new benchmark for future research.
2502.08825
Examining and Adapting Time for Multilingual Classification via Mixture of Temporal Experts
cs.CL
Time is implicitly embedded in classification process: classifiers are usually built on existing data while to be applied on future data whose distributions (e.g., label and token) may change. However, existing state-of-the-art classification models merely consider the temporal variations and primarily focus on English corpora, which leaves temporal studies less explored, let alone under multilingual settings. In this study, we fill the gap by treating time as domains (e.g., 2024 vs. 2025), examining temporal effects, and developing a domain adaptation framework to generalize classifiers over time on multiple languages. Our framework proposes Mixture of Temporal Experts (MoTE) to leverage both semantic and data distributional shifts to learn and adapt temporal trends into classification models. Our analysis shows classification performance varies over time across different languages, and we experimentally demonstrate that MoTE can enhance classifier generalizability over temporal data shifts. Our study provides analytic insights and addresses the need for time-aware models that perform robustly in multilingual scenarios.
2502.08826
Ask in Any Modality: A Comprehensive Survey on Multimodal Retrieval-Augmented Generation
cs.CL cs.AI cs.IR
Large Language Models (LLMs) struggle with hallucinations and outdated knowledge due to their reliance on static training data. Retrieval-Augmented Generation (RAG) mitigates these issues by integrating external dynamic information enhancing factual and updated grounding. Recent advances in multimodal learning have led to the development of Multimodal RAG, incorporating multiple modalities such as text, images, audio, and video to enhance the generated outputs. However, cross-modal alignment and reasoning introduce unique challenges to Multimodal RAG, distinguishing it from traditional unimodal RAG. This survey offers a structured and comprehensive analysis of Multimodal RAG systems, covering datasets, metrics, benchmarks, evaluation, methodologies, and innovations in retrieval, fusion, augmentation, and generation. We precisely review training strategies, robustness enhancements, and loss functions, while also exploring the diverse Multimodal RAG scenarios. Furthermore, we discuss open challenges and future research directions to support advancements in this evolving field. This survey lays the foundation for developing more capable and reliable AI systems that effectively leverage multimodal dynamic external knowledge bases. Resources are available at https://github.com/llm-lab-org/Multimodal-RAG-Survey.
2502.08828
A Survey on Data-Centric AI: Tabular Learning from Reinforcement Learning and Generative AI Perspective
cs.LG cs.AI
Tabular data is one of the most widely used data formats across various domains such as bioinformatics, healthcare, and marketing. As artificial intelligence moves towards a data-centric perspective, improving data quality is essential for enhancing model performance in tabular data-driven applications. This survey focuses on data-driven tabular data optimization, specifically exploring reinforcement learning (RL) and generative approaches for feature selection and feature generation as fundamental techniques for refining data spaces. Feature selection aims to identify and retain the most informative attributes, while feature generation constructs new features to better capture complex data patterns. We systematically review existing generative methods for tabular data engineering, analyzing their latest advancements, real-world applications, and respective strengths and limitations. This survey emphasizes how RL-based and generative techniques contribute to the automation and intelligence of feature engineering. Finally, we summarize the existing challenges and discuss future research directions, aiming to provide insights that drive continued innovation in this field.
2502.08829
PLayer-FL: A Principled Approach to Personalized Layer-wise Cross-Silo Federated Learning
cs.LG
Non-identically distributed data is a major challenge in Federated Learning (FL). Personalized FL tackles this by balancing local model adaptation with global model consistency. One variant, partial FL, leverages the observation that early layers learn more transferable features by federating only early layers. However, current partial FL approaches use predetermined, architecture-specific rules to select layers, limiting their applicability. We introduce Principled Layer-wise-FL (PLayer-FL), which uses a novel federation sensitivity metric to identify layers that benefit from federation. This metric, inspired by model pruning, quantifies each layer's contribution to cross-client generalization after the first training epoch, identifying a transition point in the network where the benefits of federation diminish. We first demonstrate that our federation sensitivity metric shows strong correlation with established generalization measures across diverse architectures. Next, we show that PLayer-FL outperforms existing FL algorithms on a range of tasks, also achieving more uniform performance improvements across clients.
2502.08832
LSM Trees in Adversarial Environments
cs.DB cs.CR
The Log Structured Merge (LSM) Tree is a popular choice for key-value stores that focus on optimized write throughput while maintaining performant, production-ready read latencies. To optimize read performance, LSM stores rely on a probabilistic data structure called the Bloom Filter (BF). In this paper, we focus on adversarial workloads that lead to a sharp degradation in read performance by impacting the accuracy of BFs used within the LSM store. Our evaluation shows up to $800\%$ increase in the read latency of lookups for popular LSM stores. We define adversarial models and security definitions for LSM stores. We implement adversary resilience into two popular LSM stores, LevelDB and RocksDB. We use our implementations to demonstrate how performance degradation under adversarial workloads can be mitigated.
2502.08834
A Reversible Solver for Diffusion SDEs
cs.LG cs.AI stat.ML
Diffusion models have quickly become the state-of-the-art for generation tasks across many different data modalities. An important ability of diffusion models is the ability to encode samples from the data distribution back into the sampling prior distribution. This is useful for performing alterations to real data samples along with guided generation via the continuous adjoint equations. We propose an algebraically reversible solver for diffusion SDEs that can exactly invert real data samples into the prior distribution.
2502.08835
A Bundle-based Augmented Lagrangian Framework: Algorithm, Convergence, and Primal-dual Principles
math.OC cs.SY eess.SY
We propose a new bundle-based augmented Lagrangian framework for solving constrained convex problems. Unlike the classical (inexact) augmented Lagrangian method (ALM) that has a nested double-loop structure, our framework features a $\textit{single-loop}$ process. Motivated by the proximal bundle method (PBM), we use a $\textit{bundle}$ of past iterates to approximate the subproblem in ALM to get a computationally efficient update at each iteration. We establish sub-linear convergences for primal feasibility, primal cost values, and dual iterates under mild assumptions. With further regularity conditions, such as quadratic growth, our algorithm enjoys $\textit{linear}$ convergences. Importantly, this linear convergence can happen for a class of conic optimization problems, including semidefinite programs. Our proof techniques leverage deep connections with inexact ALM and primal-dual principles with PBM.
2502.08836
Survey on Single-Image Reflection Removal using Deep Learning Techniques
cs.CV
The phenomenon of reflection is quite common in digital images, posing significant challenges for various applications such as computer vision, photography, and image processing. Traditional methods for reflection removal often struggle to achieve clean results while maintaining high fidelity and robustness, particularly in real-world scenarios. Over the past few decades, numerous deep learning-based approaches for reflection removal have emerged, yielding impressive results. In this survey, we conduct a comprehensive review of the current literature by focusing on key venues such as ICCV, ECCV, CVPR, NeurIPS, etc., as these conferences and journals have been central to advances in the field. Our review follows a structured paper selection process, and we critically assess both single-stage and two-stage deep learning methods for reflection removal. The contribution of this survey is three-fold: first, we provide a comprehensive summary of the most recent work on single-image reflection removal; second, we outline task hypotheses, current deep learning techniques, publicly available datasets, and relevant evaluation metrics; and third, we identify key challenges and opportunities in deep learning-based reflection removal, highlighting the potential of this rapidly evolving research area.
2502.08840
Thresholds for Reconstruction of Random Hypergraphs From Graph Projections
math.ST cs.IT math.IT math.PR stat.TH
The graph projection of a hypergraph is a simple graph with the same vertex set and with an edge between each pair of vertices that appear in a hyperedge. We consider the problem of reconstructing a random $d$-uniform hypergraph from its projection. Feasibility of this task depends on $d$ and the density of hyperedges in the random hypergraph. For $d=3$ we precisely determine the threshold, while for $d\geq 4$ we give bounds. All of our feasibility results are obtained by exhibiting an efficient algorithm for reconstructing the original hypergraph, while infeasibility is information-theoretic. Our results also apply to mildly inhomogeneous random hypergrahps, including hypergraph stochastic block models (HSBM). A consequence of our results is an optimal HSBM recovery algorithm, improving on a result of Guadio and Joshi in 2023.
2502.08841
Delayed takedown of illegal content on social media makes moderation ineffective
cs.SI cs.CY
Social media platforms face legal and regulatory demands to swiftly remove illegal content, sometimes under strict takedown deadlines. However, the effects of moderation speed and the impact of takedown deadlines remain underexplored. This study models the relationship between the timeliness of illegal content removal and its prevalence, reach, and exposure on social media. By simulating illegal content diffusion using empirical data from the DSA Transparency Database, we demonstrate that rapid takedown (within hours) significantly reduces illegal content prevalence and exposure, while longer delays decrease the effectiveness of moderation efforts. While these findings support tight takedown deadlines for content removal, such deadlines cannot address the delay in identifying the illegal content and can adversely affect the quality of content moderation.
2502.08844
MuJoCo Playground
cs.RO
We introduce MuJoCo Playground, a fully open-source framework for robot learning built with MJX, with the express goal of streamlining simulation, training, and sim-to-real transfer onto robots. With a simple "pip install playground", researchers can train policies in minutes on a single GPU. Playground supports diverse robotic platforms, including quadrupeds, humanoids, dexterous hands, and robotic arms, enabling zero-shot sim-to-real transfer from both state and pixel inputs. This is achieved through an integrated stack comprising a physics engine, batch renderer, and training environments. Along with video results, the entire framework is freely available at playground.mujoco.org
2502.08845
Optimal Dataset Size for Recommender Systems: Evaluating Algorithms' Performance via Downsampling
cs.IR
This thesis investigates dataset downsampling as a strategy to optimize energy efficiency in recommender systems while maintaining competitive performance. With increasing dataset sizes posing computational and environmental challenges, this study explores the trade-offs between energy efficiency and recommendation quality in Green Recommender Systems, which aim to reduce environmental impact. By applying two downsampling approaches to seven datasets, 12 algorithms, and two levels of core pruning, the research demonstrates significant reductions in runtime and carbon emissions. For example, a 30% downsampling portion can reduce runtime by 52% compared to the full dataset, leading to a carbon emission reduction of up to 51.02 KgCO2e during the training of a single algorithm on a single dataset. The analysis reveals that algorithm performance under different downsampling portions depends on factors like dataset characteristics, algorithm complexity, and the specific downsampling configuration (scenario dependent). Some algorithms, which showed lower nDCG@10 scores compared to higher-performing ones, exhibited lower sensitivity to the amount of training data, offering greater potential for efficiency in lower downsampling portions. On average, these algorithms retained 81% of full-size performance using only 50% of the training set. In certain downsampling configurations, where more users were progressively included while keeping the test set size fixed, they even showed higher nDCG@10 scores than when using the full dataset. These findings highlight the feasibility of balancing sustainability and effectiveness, providing insights for designing energy-efficient recommender systems and promoting sustainable AI practices.
2502.08856
A Systematic Evaluation of Generative Models on Tabular Transportation Data
cs.LG
The sharing of large-scale transportation data is beneficial for transportation planning and policymaking. However, it also raises significant security and privacy concerns, as the data may include identifiable personal information, such as individuals' home locations. To address these concerns, synthetic data generation based on real transportation data offers a promising solution that allows privacy protection while potentially preserving data utility. Although there are various synthetic data generation techniques, they are often not tailored to the unique characteristics of transportation data, such as the inherent structure of transportation networks formed by all trips in the datasets. In this paper, we use New York City taxi data as a case study to conduct a systematic evaluation of the performance of widely used tabular data generative models. In addition to traditional metrics such as distribution similarity, coverage, and privacy preservation, we propose a novel graph-based metric tailored specifically for transportation data. This metric evaluates the similarity between real and synthetic transportation networks, providing potentially deeper insights into their structural and functional alignment. We also introduced an improved privacy metric to address the limitations of the commonly-used one. Our experimental results reveal that existing tabular data generative models often fail to perform as consistently as claimed in the literature, particularly when applied to transportation data use cases. Furthermore, our novel graph metric reveals a significant gap between synthetic and real data. This work underscores the potential need to develop generative models specifically tailored to take advantage of the unique characteristics of emerging domains, such as transportation.
2502.08858
Estimating Probabilities of Causation with Machine Learning Models
cs.AI
Probabilities of causation play a crucial role in modern decision-making. This paper addresses the challenge of predicting probabilities of causation for subpopulations with insufficient data using machine learning models. Tian and Pearl first defined and derived tight bounds for three fundamental probabilities of causation: the probability of necessity and sufficiency (PNS), the probability of sufficiency (PS), and the probability of necessity (PN). However, estimating these probabilities requires both experimental and observational distributions specific to each subpopulation, which are often unavailable or impractical to obtain with limited population-level data. We assume that the probabilities of causation for each subpopulation are determined by its characteristics. To estimate these probabilities for subpopulations with insufficient data, we propose using machine learning models that draw insights from subpopulations with sufficient data. Our evaluation of multiple machine learning models indicates that, given sufficient population-level data and an appropriate choice of machine learning model and activation function, PNS can be effectively predicted. Through simulation studies, we show that our multilayer perceptron (MLP) model with the Mish activation function achieves a mean absolute error (MAE) of approximately 0.02 in predicting PNS for 32,768 subpopulations using data from around 2,000 subpopulations.
2502.08859
EnigmaEval: A Benchmark of Long Multimodal Reasoning Challenges
cs.AI cs.CL
As language models master existing reasoning benchmarks, we need new challenges to evaluate their cognitive frontiers. Puzzle-solving events are rich repositories of challenging multimodal problems that test a wide range of advanced reasoning and knowledge capabilities, making them a unique testbed for evaluating frontier language models. We introduce EnigmaEval, a dataset of problems and solutions derived from puzzle competitions and events that probes models' ability to perform implicit knowledge synthesis and multi-step deductive reasoning. Unlike existing reasoning and knowledge benchmarks, puzzle solving challenges models to discover hidden connections between seemingly unrelated pieces of information to uncover solution paths. The benchmark comprises 1184 puzzles of varying complexity -- each typically requiring teams of skilled solvers hours to days to complete -- with unambiguous, verifiable solutions that enable efficient evaluation. State-of-the-art language models achieve extremely low accuracy on these puzzles, even lower than other difficult benchmarks such as Humanity's Last Exam, unveiling models' shortcomings when challenged with problems requiring unstructured and lateral reasoning.
2502.08860
Brain in the Dark: Design Principles for Neuromimetic Inference under the Free Energy Principle
cs.NE
Deep learning has revolutionised artificial intelligence (AI) by enabling automatic feature extraction and function approximation from raw data. However, it faces challenges such as a lack of out-of-distribution generalisation, catastrophic forgetting and poor interpretability. In contrast, biological neural networks, such as those in the human brain, do not suffer from these issues, inspiring AI researchers to explore neuromimetic deep learning, which aims to replicate brain mechanisms within AI models. A foundational theory for this approach is the Free Energy Principle (FEP), which despite its potential, is often considered too complex to understand and implement in AI as it requires an interdisciplinary understanding across a variety of fields. This paper seeks to demystify the FEP and provide a comprehensive framework for designing neuromimetic models with human-like perception capabilities. We present a roadmap for implementing these models and a Pytorch code repository for applying FEP in a predictive coding network.
2502.08864
Off-Switching Not Guaranteed
cs.AI
Hadfield-Menell et al. (2017) propose the Off-Switch Game, a model of Human-AI cooperation in which AI agents always defer to humans because they are uncertain about our preferences. I explain two reasons why AI agents might not defer. First, AI agents might not value learning. Second, even if AI agents value learning, they might not be certain to learn our actual preferences.
2502.08866
BrainWavLM: Fine-tuning Speech Representations with Brain Responses to Language
cs.CL
Speech encoding models use auditory representations to predict how the human brain responds to spoken language stimuli. Most performant encoding models linearly map the hidden states of artificial neural networks to brain data, but this linear restriction may limit their effectiveness. In this work, we use low-rank adaptation (LoRA) to fine-tune a WavLM-based encoding model end-to-end on a brain encoding objective, producing a model we name BrainWavLM. We show that fine-tuning across all of cortex improves average encoding performance with greater stability than without LoRA. This improvement comes at the expense of low-level regions like auditory cortex (AC), but selectively fine-tuning on these areas improves performance in AC, while largely retaining gains made in the rest of cortex. Fine-tuned models generalized across subjects, indicating that they learned robust brain-like representations of the speech stimuli. Finally, by training linear probes, we showed that the brain data strengthened semantic representations in the speech model without any explicit annotations. Our results demonstrate that brain fine-tuning produces best-in-class speech encoding models, and that non-linear methods have the potential to bridge the gap between artificial and biological representations of semantics.
2502.08869
Harnessing Vision Models for Time Series Analysis: A Survey
cs.LG cs.AI cs.CV
Time series analysis has witnessed the inspiring development from traditional autoregressive models, deep learning models, to recent Transformers and Large Language Models (LLMs). Efforts in leveraging vision models for time series analysis have also been made along the way but are less visible to the community due to the predominant research on sequence modeling in this domain. However, the discrepancy between continuous time series and the discrete token space of LLMs, and the challenges in explicitly modeling the correlations of variates in multivariate time series have shifted some research attentions to the equally successful Large Vision Models (LVMs) and Vision Language Models (VLMs). To fill the blank in the existing literature, this survey discusses the advantages of vision models over LLMs in time series analysis. It provides a comprehensive and in-depth overview of the existing methods, with dual views of detailed taxonomy that answer the key research questions including how to encode time series as images and how to model the imaged time series for various tasks. Additionally, we address the challenges in the pre- and post-processing steps involved in this framework and outline future directions to further advance time series analysis with vision models.