AI Research News Feeds for October 16th, 2025

AI RESEARCH PAPERS & ACADEMIC SOURCES

• EduDial: Constructing a Large-scale Multi-turn Teacher-Student Dialogue Corpus
• Who's Asking? Evaluating LLM Robustness to Inquiry Personas in Factual Question Answering
• The Curious Case of Curiosity across Human Cultures and LLMs
• 3-Model Speculative Decoding
• A Multilingual, Large-Scale Study of the Interplay between LLM Safeguards, Personalisation, and Disinformation
• OPLoRA: Orthogonal Projection LoRA Prevents Catastrophic Forgetting during Parameter-Efficient Fine-Tuning
• On the Role of Preference Variance in Preference Optimization
• GatePro: Parameter-Free Expert Selection Optimization for Mixture-of-Experts Models
• I Am Aligned, But With Whom? MENA Values Benchmark for Evaluating Cultural Alignment and Multilingual Bias in LLMs
• Mirror Speculative Decoding: Breaking the Serial Barrier in LLM Inference
• A Matter of Representation: Towards Graph-Based Abstract Code Generation
• CoT-Evo: Evolutionary Distillation of Chain-of-Thought for Scientific Reasoning
• Putting on the Thinking Hats: A Survey on Chain of Thought Fine-tuning from the Perspective of Human Reasoning Mechanism
• DSCD: Large Language Model Detoxification with Self-Constrained Decoding
• SHIELD: Classifier-Guided Prompting for Robust and Safer LVLMs
• Grounding Long-Context Reasoning with Contextual Normalization for Retrieval-Augmented Generation
• Text Anomaly Detection with Simplified Isolation Kernel
• A fully automated and scalable Parallel Data Augmentation for Low Resource Languages using Image and Text Analytics
• Hierarchical Frequency Tagging Probe (HFTP): A Unified Approach to Investigate Syntactic Structure Representations in Large Language Models and the Human Brain
• Do You Get the Hint? Benchmarking LLMs on the Board Game Concept
• Beyond Correctness: Rewarding Faithful Reasoning in Retrieval-Augmented Generation
• In-Distribution Steering: Balancing Control and Coherence in Language Model Generation
• Mismatch Aware Guidance for Robust Emotion Control in Auto-Regressive TTS Models
• ChatR1: Reinforcement Learning for Conversational Reasoning and Retrieval Augmented Question Answering
• Embedding-Based Context-Aware Reranker
• Taming the Fragility of KV Cache Eviction in LLM Inference
• Are Proverbs the New Pythian Oracles? Exploring Sentiment in Greek Sayings
• D-SMART: Enhancing LLM Dialogue Consistency via Dynamic Structured Memory And Reasoning Tree
• Make an Offer They Can't Refuse: Grounding Bayesian Persuasion in Real-World Dialogues without Pre-Commitment
• Doing Things with Words: Rethinking Theory of Mind Simulation in Large Language Models
• Investigating Lexical Change through Cross-Linguistic Colexification Patterns
• Evaluating Arabic Large Language Models: A Survey of Benchmarks, Methods, and Gaps
• Beyond Single-Reward: Multi-Pair, Multi-Perspective Preference Optimization for Machine Translation
• Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization
• Sparse Subnetwork Enhancement for Underrepresented Languages in Large Language Models
• FreshTab: Sourcing Fresh Data for Table-to-Text Generation Evaluation
• MemoTime: Memory-Augmented Temporal Knowledge Graph Enhanced Large Language Model Reasoning
• How Sampling Affects the Detectability of Machine-written texts: A Comprehensive Study
• GAPS: A Clinically Grounded, Automated Benchmark for Evaluating AI Clinicians
• Assessing Web Search Credibility and Response Groundedness in Chat Assistants
• Confidence-Based Response Abstinence: Improving LLM Trustworthiness via Activation-Based Uncertainty Estimation
• The Mechanistic Emergence of Symbol Grounding in Language Models
• Breadcrumbs Reasoning: Memory-Efficient Reasoning with Compression Beacons
• BRIEF-Pro: Universal Context Compression with Short-to-Long Synthesis for Fast and Accurate Multi-Hop Reasoning
• Toward LLM-Supported Automated Assessment of Critical Thinking Subskills
• Unifying Vision-Language Latents for Zero-label Image Caption Enhancement
• UNCAP: Uncertainty-Guided Planning Using Natural Language Communication for Cooperative Autonomous Vehicles
• Addressing the alignment problem in transportation policy making: an LLM approach
• MMLongCite: A Benchmark for Evaluating Fidelity of Long-Context Vision-Language Models
• Two Heads Are Better Than One: Audio-Visual Speech Error Correction with Dual Hypotheses
• UniMoE-Audio: Unified Speech and Music Generation with Dynamic-Capacity MoE
• LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models
• Towards Region-aware Bias Evaluation Metrics
• ICA-RAG: Information Completeness Guided Adaptive Retrieval-Augmented Generation for Disease Diagnosis
• Teaching Models to Understand (but not Generate) High-risk Data
• What Does Neuro Mean to Cardio? Investigating the Role of Clinical Specialty Data in Medical LLMs
• RPM: Reasoning-Level Personalization for Black-Box Large Language Models
• RedTeamCUA: Realistic Adversarial Testing of Computer-Use Agents in Hybrid Web-OS Environments
• A Linguistically Motivated Analysis of Intonational Phrasing in Text-to-Speech Systems: Revealing Gaps in Syntactic Sensitivity
• The Landscape of Arabic Large Language Models (ALLMs): A New Era for Arabic Language Technology
• MMD-Flagger: Leveraging Maximum Mean Discrepancy to Detect Hallucinations
• SemVink: Advancing VLMs' Semantic Understanding of Optical Illusions via Visual Global Thinking
• KG2QA: Knowledge Graph-enhanced Retrieval-augmented Generation for Communication Standards Question Answering
• Aligning Large Language Models to Low-Resource Languages through LLM-Based Selective Translation: A Systematic Study
• Assessing the Latent Automated Program Repair Capabilities of Large Language Models using Round-Trip Translation
• Rec-R1: Bridging Generative Large Language Models and User-Centric Recommendation Systems via Reinforcement Learning
• Benchmarking LLMs' Swarm intelligence
• MERIT: Multilingual Semantic Retrieval with Interleaved Multi-Condition Query
• Cross-modal Associations in Vision and Language Models: Revisiting the Bouba-Kiki Effect
• Learning to Explore in Diverse Reward Settings via Temporal-Difference-Error Maximization
• HEAL: An Empirical Study on Hallucinations in Embodied Agents Driven by Large Language Models
• Riemannian generative decoder
• Pad\'e Approximant Neural Networks for Enhanced Electric Motor Fault Diagnosis Using Vibration and Acoustic Data
• Context-Action Embedding Learning for Off-Policy Evaluation in Contextual Bandits
• PPA-Game: Characterizing and Learning Competitive Dynamics Among Online Content Creators
• Learning to sample fibers for goodness-of-fit testing
• Beyond the noise: intrinsic dimension estimation with optimal neighbourhood identification
• Quantum Circuit Synthesis and Compilation Optimization: Overview and Prospects
• Persistent Homology via Ellipsoids
• Computing Systemic Risk Measures with Graph Neural Networks
• LLMBridge: Reducing Costs in a Prompt-Centric Internet
• Enhancing the reliability of machine learning for gravitational wave parameter estimation with attention-based models
• A Statistical Theory of Contrastive Learning via Approximate Sufficient Statistics
• Gen-C: Populating Virtual Worlds with Generative Crowds
• BrainOmni: A Brain Foundation Model for Unified EEG and MEG Signals
• A Unified Framework for Variable Selection in Model-Based Clustering with Missing Not at Random
• Physics-Based Machine Learning Closures and Wall Models for Hypersonic Transition-Continuum Boundary Layer Predictions
• Efficient Restarts in Non-Stationary Model-Free Reinforcement Learning
• On efficiently computable functions, deep networks and sparse compositionality
• QLENS: Towards A Quantum Perspective of Language Transformers
• Learning by Steering the Neural Dynamics: A Statistical Mechanics Perspective
• Nonlinear discretizations and Newton's method: characterizing stationary points of regression objectives
• Mamba Can Learn Low-Dimensional Targets In-Context via Test-Time Feature Learning
• Influence Dynamics and Stagewise Data Attribution
• GraphShaper: Geometry-aware Alignment for Improving Transfer Learning in Text-Attributed Graphs
• H4G: Unlocking Faithful Inference for Zero-Shot Graph Learning in Hyperbolic Space
• Rethinking the Role of Dynamic Sparse Training for Scalable Deep Reinforcement Learning
• nuGPR: GPU-Accelerated Gaussian Process Regression with Iterative Algorithms and Low-Rank Approximations
• Graph Few-Shot Learning via Adaptive Spectrum Experts and Cross-Set Distribution Calibration
• Fairness-Constrained Optimization Attack in Federated Learning
• Self-Verifying Reflection Helps Transformers with CoT Reasoning
• Hierarchical Koopman Diffusion: Fast Generation with Interpretable Diffusion Trajectory
• Unveiling the Vulnerability of Graph-LLMs: An Interpretable Multi-Dimensional Adversarial Attack on TAGs
• Optimal Regularization for Performative Learning
• FedMMKT:Co-Enhancing a Server Text-to-Image Model and Client Task Models in Multi-Modal Federated Learning
• Multi-Action Self-Improvement for Neural Combinatorial Optimization
• General Fourier Feature Physics-Informed Extreme Learning Machine (GFF-PIELM) for High-Frequency PDEs
• Leveraging Teleconnections with Physics-Informed Graph Attention Networks for Long-Range Extreme Rainfall Forecasting in Thailand
• Traveling Salesman-Based Token Ordering Improves Stability in Homomorphically Encrypted Language Models
• Towards Cross-Modal Error Detection with Tables and Images
• Enhanced Pre-training of Graph Neural Networks for Million-Scale Heterogeneous Graphs
• Cautious Weight Decay
• Continuous Uniqueness and Novelty Metrics for Generative Modeling of Inorganic Crystals
• Bayesian Optimization for Dynamic Pricing and Learning
• Time-Correlated Video Bridge Matching
• CrossAD: Time Series Anomaly Detection with Cross-scale Associations and Cross-window Modeling
• Mitigating the Noise Shift for Denoising Generative Models via Noise Awareness Guidance
• Multi-Armed Bandits with Minimum Aggregated Revenue Constraints
• Research in Collaborative Learning Does Not Serve Cross-Silo Federated Learning in Practice
• Towards Fast Coarse-graining and Equation Discovery with Foundation Inference Models
• Expert or not? assessing data quality in offline reinforcement learning
• On Foundation Models for Temporal Point Processes to Accelerate Scientific Discovery
• Towards Foundation Inference Models that Learn ODEs In-Context
• Structured Sparsity and Weight-adaptive Pruning for Memory and Compute efficient Whisper models
• Structure-Aware Spectral Sparsification via Uniform Edge Sampling
• Keep Calm and Avoid Harmful Content: Concept Alignment and Latent Manipulation Towards Safer Answers
• CoRA: Covariate-Aware Adaptation of Time Series Foundation Models
• Few Shot Semi-Supervised Learning for Abnormal Stop Detection from Sparse GPS Trajectories
• Multitask finetuning and acceleration of chemical pretrained models for small molecule drug property prediction
• CARVQ: Corrective Adaptor with Group Residual Vector Quantization for LLM Embedding Compression
• Improving Decision Trees through the Lens of Parameterized Local Search
• Doctor Rashomon and the UNIVERSE of Madness: Variable Importance with Unobserved Confounding and the Rashomon Effect
• KoALA: KL-L0 Adversarial Detector via Label Agreement
• Sample-Efficient Omniprediction for Proper Losses
• scPPDM: A Diffusion Model for Single-Cell Drug-Response Prediction
• Multi-objective Bayesian Optimization with Human-in-the-Loop for Flexible Neuromorphic Electronics Fabrication
• Quantum Kernel Methods: Convergence Theory, Separation Bounds and Applications to Marketing Analytics
• PRISM: Enhancing Protein Inverse Folding through Fine-Grained Retrieval on Structure-Sequence Multimodal Representations
• Dimension-Free Minimax Rates for Learning Pairwise Interactions in Attention-Style Models
• On Thompson Sampling and Bilateral Uncertainty in Additive Bayesian Optimization
• Active Subspaces in Infinite Dimension
• High-Probability Bounds For Heterogeneous Local Differential Privacy
• Simplifying Optimal Transport through Schatten-$p$ Regularization
• Enhancing Diffusion-Based Sampling with Molecular Collective Variables
• Statistical Guarantees for High-Dimensional Stochastic Gradient Descent
• Embedding the Teacher: Distilling vLLM Preferences for Scalable Image Retrieval
• MIARec: Mutual-influence-aware Heterogeneous Network Embedding for Scientific Paper Recommendation
• Compressibility Measures Complexity: Minimum Description Length Meets Singular Learning Theory
• FedLoDrop: Federated LoRA with Dropout for Generalized LLM Fine-tuning
• Locket: Robust Feature-Locking Technique for Language Models
• Probabilistic Super-Resolution for Urban Micrometeorology via a Schr\"odinger Bridge
• Follow-the-Perturbed-Leader for Decoupled Bandits: Best-of-Both-Worlds and Practicality
• Learning Mean-Field Games through Mean-Field Actor-Critic Flow
• Controllable Collision Scenario Generation via Collision Pattern Prediction
• A Gradient Guided Diffusion Framework for Chance Constrained Programming
• The Living Forecast: Evolving Day-Ahead Predictions into Intraday Reality
• Heterogeneous RBCs via deep multi-agent reinforcement learning
• DeepTrust: Multi-Step Classification through Dissimilar Adversarial Representations for Robust Android Malware Detection
• Learning Latent Energy-Based Models via Interacting Particle Langevin Dynamics
• Pretraining in Actor-Critic Reinforcement Learning for Robot Motion Control
• Constrained Sensing and Reliable State Estimation with Shallow Recurrent Decoders on a TRIGA Mark II Reactor
• Improved Central Limit Theorem and Bootstrap Approximations for Linear Stochastic Approximation
• Improving Generative Behavior Cloning via Self-Guidance and Adaptive Chunking
• Robot Learning: A Tutorial
• Geopolitics, Geoeconomics and Risk:A Machine Learning Approach
• Neural Guided Sampling for Quantum Circuit Optimization
• Formal Models and Convergence Analysis for Context-Aware Security Verification
• Diff-XYZ: A Benchmark for Evaluating Diff Understanding
• Why the noise model matters: A performance gap in learned regularization
• Universal Adaptive Environment Discovery
• Same model, better performance: the impact of shuffling on DNA Language Models benchmarking
• Adapting Noise to Data: Generative Flows from 1D Processes
• Contraction and entropy production in continuous-time Sinkhorn dynamics
• Data-Model Co-Evolution: Growing Test Sets to Refine LLM Behavior
• Dendrograms of Mixing Measures for Softmax-Gated Gaussian Mixture of Experts: Consistency without Model Sweeps
• Wavefront Coding for Accommodation-Invariant Near-Eye Displays
• Posterior Sampling for Continuing Environments
• WW-FL: Secure and Private Large-Scale Federated Learning
• IBCL: Zero-shot Model Generation under Stability-Plasticity Trade-offs
• Competitive Advantage Attacks to Decentralized Federated Learning
• Optimistic Multi-Agent Policy Gradient
• Toward Fair Graph Neural Networks Via Dual-Teacher Knowledge Distillation
• AlphaZero Neural Scaling and Zipf's Law: a Tale of Board Games and Power Laws
• Dimension Reduction with Locally Adjusted Graphs
• Resource-Constrained Federated Continual Learning: What Does Matter?
• Evaluating multiple models using labeled and unlabeled data
• RIGNO: A Graph-based framework for robust and accurate operator learning for PDEs on arbitrary domains
• Mirror Descent Actor Critic via Bounded Advantage Learning
• Provably Efficient RL under Episode-Wise Safety in Constrained MDPs with Linear Function Approximation
• Wasserstein-based Kernel Principal Component Analysis for Clustering Applications
• MooseAgent: A LLM Based Multi-agent Framework for Automating Moose Simulation
• Generative Deep Learning Framework for Inverse Design of Fuels
• Newton-Puiseux Analysis for Interpretability and Calibration of Complex-Valued Neural Networks
• Variational Rank Reduction Autoencoders
• Panda: A pretrained forecast model for chaotic dynamics
• Narrow Operator Models of Stellarator Equilibria in Fourier Zernike Basis
• K-Merge: Online Continual Merging of Adapters for On-device Large Language Models
• In-Browser LLM-Guided Fuzzing for Real-Time Prompt Injection Testing in Agentic AI Browsers
• Modeling Cultural Bias in Facial Expression Recognition with Adaptive Agents
• OpenDerisk: An Industrial Framework for AI-Driven SRE, with Design, Implementation, and Case Studies
• Deflanderization for Game Dialogue: Balancing Character Authenticity with Task Execution in LLM-based NPCs
• Subject Roles in the EU AI Act: Mapping and Regulatory Implications
• NOSA: Native and Offloadable Sparse Attention
• Message Passing on the Edge: Towards Scalable and Expressive GNNs
• The Role of Computing Resources in Publishing Foundation Model Research
• Unlocking Public Catalogues: Instruction-Tuning LLMs for ICD Coding of German Tumor Diagnoses
• Closing the Gap Between Text and Speech Understanding in LLMs
• Time Series Foundation Models: Benchmarking Challenges and Requirements
• Axial Neural Networks for Dimension-Free Foundation Models
• CanvasMAR: Improving Masked Autoregressive Video Generation With Canvas
• MVCustom: Multi-View Customized Diffusion via Geometric Latent Rendering and Completion
• Simplicial Embeddings Improve Sample Efficiency in Actor-Critic Agents
• Dedelayed: Deleting remote inference delay via on-device correction
• NExT-OMNI: Towards Any-to-Any Omnimodal Foundation Models with Discrete Flow Matching
• FIRST: Federated Inference Resource Scheduling Toolkit for Scientific AI Model Access
• Multi-Scale High-Resolution Logarithmic Grapher Module for Efficient Vision GNNs
• RECODE: Reasoning Through Code Generation for Visual Question Answering
• Scaling Vision Transformers for Functional MRI with Flat Maps
• InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy
• The Art of Scaling Reinforcement Learning Compute for LLMs
• Provably Invincible Adversarial Attacks on Reinforcement Learning Systems: A Rate-Distortion Information-Theoretic Approach
• Bee: A High-Quality Corpus and Full-Stack Suite to Unlock Advanced Fully Open MLLMs
• Generative Universal Verifier as Multimodal Meta-Reasoner
• Quantile Markov Decision Process
• Translating Regulatory Clauses into Executable Codes for Building Design Checking via Large Language Model Driven Function Matching and Composing
• Improving Planning with Large Language Models: A Modular Agentic Architecture
• Sentiment and Emotion-aware Multi-criteria Fuzzy Group Decision Making System
• Reinforcing Competitive Multi-Agents for Playing 'So Long Sucker'
• AI Realtor: Towards Grounded Persuasive Language Generation for Automated Copywriting
• LLM-Enabled In-Context Learning for Data Collection Scheduling in UAV-assisted Sensor Networks
• Deep Generative Prior for First Order Inverse Optimization
• GUARDIAN: Safeguarding LLM Multi-Agent Collaborations with Temporal Graph Modeling
• MSEarth: A Multimodal Scientific Dataset and Benchmark for Phenomena Uncovering in Earth Science
• Convergence and sample complexity of natural policy gradient primal-dual methods for constrained MDPs
• When to Trust Your Simulator: Dynamics-Aware Hybrid Offline-and-Online Reinforcement Learning
• Nash Equilibria, Regularization and Computation in Optimal Transport-Based Distributionally Robust Optimization
• MULTI: Multimodal Understanding Leaderboard with Text and Images
• Do LLM Agents Have Regret? A Case Study in Online Learning and Games
• A Comprehensive Survey on Data Augmentation
• Extreme Compression of Adaptive Neural Images
• Beyond Visual Appearances: Privacy-sensitive Objects Identification via Hybrid Graph Reasoning
• Hi-Drive: Hierarchical POMDP Planning for Safe Autonomous Driving in Diverse Urban Environments
• Temporal-Difference Variational Continual Learning
• Optimal Quantization for Matrix Multiplication
• ProReason: Multi-Modal Proactive Reasoning with Decoupled Eyesight and Wisdom
• On the Limits of Language Generation: Trade-Offs Between Hallucination and Mode Collapse
• Semantically Guided Action Anticipation
• SoundnessBench: A Soundness Benchmark for Neural Network Verifiers
• BoxingGym: Benchmarking Progress in Automated Experimental Design and Model Discovery
• Position: The Artificial Intelligence and Machine Learning Community Should Adopt a More Transparent and Regulated Peer Review Process
• FALCON: Fine-grained Activation Manipulation by Contrastive Orthogonal Unalignment for Large Language Model
• PRISM: Self-Pruning Intrinsic Selection Method for Training-Free Multimodal Data Selection
• Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding
• Universal Speech Token Learning via Low-Bitrate Neural Codec and Pretrained Representations
• A Personalized Data-Driven Generative Model of Human Repetitive Motion
• On the Consistency of Multilingual Context Utilization in Retrieval-Augmented Generation
• Statistical post-processing yields accurate probabilistic forecasts from Artificial Intelligence weather models
• MIRROR: Multimodal Cognitive Reframing Therapy for Rolling with Resistance
• FineScope : Precision Pruning for Domain-Specialized Large Language Models Using SAE-Guided Self-Data Cultivation
• Flattening Hierarchies with Policy Bootstrapping
• Fact-R1: Towards Explainable Video Misinformation Detection with Deep Reasoning
• R$^2$ec: Towards Large Recommender Models with Reasoning
• ReasoningShield: Safety Detection over Reasoning Traces of Large Reasoning Models
• Multi-Scale Probabilistic Generation Theory: A Unified Information-Theoretic Framework for Hierarchical Structure in Large Language Models
• Reinforcement Learning for Out-of-Distribution Reasoning in LLMs: An Empirical Study on Diagnosis-Related Group Coding
• The quest for the GRAph Level autoEncoder (GRALE)
• FLEX: A Largescale Multimodal, Multiview Dataset for Learning Structured Representations for Fitness Action Quality Assessment
• Decentralizing Multi-Agent Reinforcement Learning with Temporal Causal Information
• Self-Predictive Representations for Combinatorial Generalization in Behavioral Cloning
• PAL: Probing Audio Encoders via LLMs - Audio Information Transfer into LLMs
• Investigating the interaction of linguistic and mathematical reasoning in language models using multilingual number puzzles
• A Brain-to-Population Graph Learning Framework for Diagnosing Brain Disorders
• LLM Probability Concentration: How Alignment Shrinks the Generative Horizon
• Orthogonal Finetuning Made Scalable
• Early Signs of Steganographic Capabilities in Frontier LLMs
• Adversarial Distilled Retrieval-Augmented Guarding Model for Online Malicious Intent Detection
• Think as a Doctor: An Interpretable AI Approach for ICU Mortality Prediction
• Schr\"odinger bridge for generative AI: Soft-constrained formulation and convergence analysis
• Z0-Inf: Zeroth Order Approximation for Data Influence
• WaveletDiff: Multilevel Wavelet Diffusion For Time Series Generation
• Evaluating Open-Source Vision-Language Models for Multimodal Sarcasm Detection
• Actor-Enriched Time Series Forecasting of Process Performance
• Improving Knowledge Graph Embeddings through Contrastive Learning with Negative Statements
• Robust Adversarial Reinforcement Learning in Stochastic Games via Sequence Modeling
• ADARL: Adaptive Low-Rank Structures for Robust Policy Learning under Uncertainty
• Variational Mixture of Graph Neural Experts for Alzheimer's Disease Biomarker Recognition in EEG Brain Networks
• From Literal to Liberal: A Meta-Prompting Framework for Eliciting Human-Aligned Exception Handling in Large Language Models
• DeepPlanner: Scaling Planning Capability for Deep Research Agents via Advantage Shaping
• SENTINEL: A Multi-Level Formal Framework for Safety Evaluation of LLM-based Embodied Agents
• From Narratives to Probabilistic Reasoning: Predicting and Interpreting Drivers' Hazardous Actions in Crashes Using Large Language Model
• Toward Reasoning-Centric Time-Series Analysis
• Repairing Reward Functions with Human Feedback to Mitigate Reward Hacking
• Emotional Cognitive Modeling Framework with Desire-Driven Objective Optimization for LLM-empowered Agent in Social Simulation
• Adaptive Reasoning Executor: A Collaborative Agent System for Efficient Reasoning
• Personalized Learning Path Planning with Goal-Driven Learner State Modeling
• EvoTest: Evolutionary Test-Time Learning for Self-Improving Agentic Systems
• An Analytical Framework to Enhance Autonomous Vehicle Perception for Smart Cities
• SAJA: A State-Action Joint Attack Framework on Multi-Agent Deep Reinforcement Learning
• Learnable Game-theoretic Policy Optimization for Data-centric Self-explanation Rationalization
• Assessing LLM Reasoning Through Implicit Causal Chain Discovery in Climate Discourse
• Mobile Coverage Analysis using Crowdsourced Data
• Confidence as a Reward: Transforming LLMs into Reward Models
• A Methodology for Assessing the Risk of Metric Failure in LLMs Within the Financial Domain
• Tandem Training for Language Models
• A Modal Logic for Temporal and Jurisdictional Classifier Models
• Training LLM Agents to Empower Humans
• From Refusal to Recovery: A Control-Theoretic Approach to Generative AI Guardrails
• Hard2Verify: A Step-Level Verification Benchmark for Open-Ended Frontier Math
• AutoCode: LLMs as Problem Setters for Competitive Programming
• Benchmarking Open-Source Large Language Models for Persian in Zero-Shot and Few-Shot Learning
• Cancer Diagnosis Categorization in Electronic Health Records Using Large Language Models and BioBERT: Model Performance Evaluation Study
• From Noise to Signal to Selbstzweck: Reframing Human Label Variation in the Era of Post-training in NLP
• MEDEQUALQA: Evaluating Biases in LLMs with Counterfactual Reasoning
• Beyond Discrete Categories: Multi-Task Valence-Arousal Modeling for Pet Vocalization Analysis
• Evidence Without Injustice: A New Counterfactual Test for Fair Algorithms
• Classifier-Augmented Generation for Structured Workflow Prediction
• Scheming Ability in LLM-to-LLM Strategic Interactions
• Automatic Speech Recognition in the Modern Era: Architectures, Training, and Evaluation
• Mathematics with large language models as provers and verifiers
• Gobernanza y trazabilidad "a prueba de AI Act" para casos de uso legales: un marco t\'ecnico-jur\'idico, m\'etricas forenses y evidencias auditables
• MTSQL-R1: Towards Long-Horizon Multi-Turn Text-to-SQL via Agentic Training
• Coherent Load Profile Synthesis with Conditional Diffusion for LV Distribution Network Scenario Generation
• Gelina: Unified Speech and Gesture Synthesis via Interleaved Token Prediction
• Repurposing Annotation Guidelines to Instruct LLM Annotators: A Case Study
• Semantic knowledge guides innovation and drives cultural evolution
• A\textsuperscript{2}FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning
• FaStFACT: Faster, Stronger Long-Form Factuality Evaluations in LLMs
• VLURes: Benchmarking VLM Visual and Linguistic Understanding in Low-Resource Languages
• Ethic-BERT: An Enhanced Deep Learning Model for Ethical and Non-Ethical Content Classification
• Efficient Adaptive Transformer: An Empirical Study and Reproducible Framework
• Adaptive Generation of Bias-Eliciting Questions for LLMs
• A Critical Review of the Need for Knowledge-Centric Evaluation of Quranic Recitation
• Three Lenses on the AI Revolution: Risk, Transformation, Continuity
• KVCOMM: Online Cross-context KV-cache Communication for Efficient LLM-based Multi-agent Systems
• InferA: A Smart Assistant for Cosmological Ensemble Data
• HyWA: Hypernetwork Weight Adapting Personalized Voice Activity Detection
• SpareCodeSearch: Searching for Code Context When You Have No Spare GPU
• Epistemic-aware Vision-Language Foundation Model for Fetal Ultrasound Interpretation
• A Multimodal XAI Framework for Trustworthy CNNs and Bias Detection in Deep Representation Learning
• Max It or Miss It: Benchmarking LLM On Solving Extremal Problems
• CurLL: A Developmental Framework to Evaluate Continual Learning in Language Models
• Developing and Validating the Arabic Version of the Attitudes Toward Large Language Models Scale
• Deliberate Lab: A Platform for Real-Time Human-AI Social Experiments
• Randomness and Interpolation Improve Gradient Descent
• SeqBench: Benchmarking Sequential Narrative Generation in Text-to-Video Models
• SceneAdapt: Scene-aware Adaptation of Human Motion Diffusion
• Time-Varying Optimization for Streaming Data Via Temporal Weighting
• VLA-0: Building State-of-the-Art VLAs with Zero Modification
• Towards Human-Centric Intelligent Treatment Planning for Radiation Therapy
• True Self-Supervised Novel View Synthesis is Transferable
• NeuroRVQ: Multi-Scale EEG Tokenization for Generative Large Brainwave Models
• Transformer-based Scalable Beamforming Optimization via Deep Residual Learning
• Agentic Discovery: Closing the Loop with Cooperative Agents
• A Multi-dimensional Semantic Surprise Framework Based on Low-Entropy Semantic Manifolds for Fine-Grained Out-of-Distribution Detection
• ESI: Epistemic Uncertainty Quantification via Semantic-preserving Intervention for Large Language Models
• TRUSTVIS: A Multi-Dimensional Trustworthiness Evaluation Framework for Large Language Models
• DriveCritic: Towards Context-Aware, Human-Aligned Evaluation for Autonomous Driving with Vision-Language Models
• Multi-Label Clinical Text Eligibility Classification and Summarization System
• On the Reasoning Abilities of Masked Diffusion Language Models
• Stable LLM Ensemble: Interaction between Example Representativeness and Diversity
• Program of Thoughts for Financial Reasoning: Leveraging Dynamic In-Context Examples and Generative Retrieval
• Behavioral Embeddings of Programs: A Quasi-Dynamic Approach for Optimization Prediction
• StressTransfer: Stress-Aware Speech-to-Speech Translation with Emphasis Preservation
• Paper Copilot: Tracking the Evolution of Peer Review in AI Conferences
• LLM-Guided Synthetic Augmentation (LGSA) for Mitigating Bias in AI Systems
• CleverCatch: A Knowledge-Guided Weak Supervision Model for Fraud Detection
• MimicParts: Part-aware Style Injection for Speech-Driven 3D Motion Generation
• What "Not" to Detect: Negation-Aware VLMs via Structured Reasoning and Token Merging
• MotionBeat: Motion-Aligned Music Representation via Embodied Contrastive Learning and Bar-Equivariant Contact-Aware Encoding
• Real-Time Crowd Counting for Embedded Systems with Lightweight Architecture
• A Ratio-Based Shapley Value for Collaborative Machine Learning - Extended Version
• To Steer or Not to Steer? Mechanistic Error Reduction with Abstention for Language Models
• Higher Satisfaction, Lower Cost: A Technical Report on How LLMs Revolutionize Meituan's Intelligent Interaction Systems
• LLM one-shot style transfer for Authorship Attribution and Verification
• Self-Augmented Visual Contrastive Decoding
• Injection, Attack and Erasure: Revocable Backdoor Attacks via Machine Unlearning
• Thompson Sampling via Fine-Tuning of LLMs
• AOAD-MAT: Transformer-based multi-agent deep reinforcement learning model considering agents' order of action decisions
• Protect: Towards Robust Guardrailing Stack for Trustworthy Enterprise LLM Systems
• Personal Attribute Leakage in Federated Speech Models
• Adversarial Fine-tuning in Offline-to-Online Reinforcement Learning for Robust Robot Control
• Generalist++: A Meta-learning Framework for Mitigating Trade-off in Adversarial Training
• Language as a Label: Zero-Shot Multimodal Classification of Everyday Postures under Data Scarcity
• Document Intelligence in the Era of Large Language Models: A Survey
• A New Perspective on Transformers in Online Reinforcement Learning for Continuous Control
• MADREC: A Multi-Aspect Driven LLM Agent for Explainable and Adaptive Recommendation
• From Minimal Existence to Human Definition: The CES-IMU-HSG Theoretical Framework
• Semantic Communication Enabled Holographic Video Processing and Transmission
• Rectify and Align GPS Points to Parking Spots via Rank-1 Constraint
• Neural Sum-of-Squares: Certifying the Nonnegativity of Polynomials with Transformers
• LiteraryQA: Towards Effective Evaluation of Long-document Narrative QA
• DistilCLIP-EEG: Enhancing Epileptic Seizure Detection Through Multi-modal Learning and Knowledge Distillation
• ConsintBench: Evaluating Language Models on Real-World Consumer Intent Understanding
• MedREK: Retrieval-Based Editing for Medical LLMs with Key-Aware Prompts
• Offline and Online KL-Regularized RLHF under Differential Privacy
• UniME-V2: MLLM-as-a-Judge for Universal Multimodal Embedding Learning