Papers

New on arXiv

Wed 5 Nov 2025

290 papers

The Illusion of Procedural Reasoning: Measuring Long-Horizon FSM Execution in LLMsLLM reasoning → Reasoning benchmarks and evaluation
Principled Coarse-Grained Acceptance for Speculative Decoding in SpeechEfficient LLM inference and serving → Speculative decoding
Emotion Recognition in Multi-Speaker Conversations through Speaker Identification, Knowledge Distillation, and Hierarchical FusionKnowledge distillation → Multimodal distillation
AnchorTP: Resilient LLM Inference with State-Preserving Elastic Tensor ParallelismMixture-of-experts → Efficient MoE training and inference systems
Diffusion-Based Image Editing: An Unforeseen Adversary to Robust Invisible WatermarksDiffusion models → Robustness, privacy and adversarial aspects +1
CoPRIS: Efficient and Stable Reinforcement Learning via Concurrency-Controlled Partial Rollout with Importance SamplingDeep reinforcement learning methods → RL training of language and agent models
SWAP: Towards Copyright Auditing of Soft Prompts via Sequential WatermarkingMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
Jailbreaking in the HaystackLLM agents and tool use → Agent threats, guardrails and governance +1
Inference-Time Optimization of Prompt Embeddings in Diffusion Models: A Comparison of sep-CMA-ES and AdamDiffusion models → Image generation, text-to-image and editing
Context informs pragmatic interpretation in vision-language modelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Investigating Robot Control Policy Learning for Autonomous X-ray-guided Spine ProceduresOffline RL and imitation learning from data → Safe offline RL
Evaluating Machine Translation Datasets for Low-Web Data Languages: A Gendered LensLLM safety and security → Safety alignment, refusal and guardrails
KnowThyself: An Agentic Assistant for LLM InterpretabilityLLM agents and tool use → General LLM agent frameworks and benchmarks
To See or To Read: User Behavior Reasoning in Multimodal LLMsLLM reasoning → Multimodal reasoning +2
ASAP: an Agentic Solution to Auto-optimize Performance of Large-Scale LLM TrainingLLM agents and tool use → General LLM agent frameworks and benchmarks
Security Analysis of Agentic AI Communication Protocols: A Comparative EvaluationLLM agents and tool use → Agent threats, guardrails and governance
Enhancing Q-Value Updates in Deep Q-Learning via Successor-State PredictionDeep reinforcement learning methods → Deep RL algorithms and representations
From Static Constraints to Dynamic Adaptation: Sample-Level Constraint Relaxation for Offline-to-Online Reinforcement LearningOffline RL and imitation learning from data → Offline RL algorithms
STARS: Synchronous Token Alignment for Robust Supervision in Large Language ModelsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Optimizing Reasoning Efficiency through Prompt Difficulty PredictionLLM reasoning → Test-time compute and efficient thinking
Expert Evaluation of LLM World Models: A High-Tc Superconductivity Case StudyWorld models → World models for language agents
Contamination Detection for VLMs using Multi-Modal Semantic PerturbationMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
What's in Common? Multimodal Models Hallucinate When Reasoning Across ScenesLLM reasoning → Multimodal reasoning +1
Leveraging LLM-based agents for social science research: insights from citation network simulationsLLM agents and tool use → Social simulation and multi-agent debate
Outbidding and Outbluffing Elite Humans: Mastering Liar's Poker via Self-Play and Reinforcement LearningDeep reinforcement learning methods → RL for games and combinatorial optimization
Shrinking the Variance: Shrinkage Baselines for Reinforcement Learning with Verifiable RewardsLLM reasoning → Reinforcement learning for reasoning
AnaFlow: Agentic LLM-based Workflow for Reasoning-Driven Explainable and Sample-Efficient Analog Circuit SizingLLM reasoning → Reasoning for agents, planning and code +1
Behavior-Adaptive Q-Learning: A Unifying Framework for Offline-to-Online RLOffline RL and imitation learning from data → Offline RL algorithms
The OpenHands Software Agent SDK: A Composable and Extensible Foundation for Production AgentsLLM agents and tool use → Software engineering coding agents
Watermarking Large Language Models in Europe: Interpreting the AI Act in Light of TechnologyLLM safety and security → Watermarking generated content and models
LiveTradeBench: Seeking Real-World Alpha with Large Language ModelsLLM reasoning → Reasoning for agents, planning and code
Towards Formalizing Reinforcement Learning Theory: A Robbins-Siegmund ApproachDeep reinforcement learning methods → RL theory and sample complexity
Going Beyond Expert Performance via Deep Implicit Imitation Reinforcement LearningOffline RL and imitation learning from data → Imitation learning methods +1
Tensor-Efficient High-Dimensional Q-learningDeep reinforcement learning methods → RL theory and sample complexity
Imitation Learning in the Deep Learning Era: A Novel Taxonomy and Recent AdvancesOffline RL and imitation learning from data → Imitation learning methods
ASVRI-Legal: Fine-Tuning LLMs with Retrieval Augmented Generation for Enhanced Legal RegulationRetrieval-augmented generation → Legal and financial RAG
MultiZebraLogic: A Multilingual Logical Reasoning BenchmarkLLM reasoning → Reasoning benchmarks and evaluation
Learning Without Critics? Revisiting GRPO in Classical Reinforcement Learning EnvironmentsLLM reasoning → Reinforcement learning for reasoning
U2F: Encouraging SWE-Agent to Seize Novelty without Losing FeasibilityLLM agents and tool use → Software engineering coding agents
HaluMem: Evaluating Hallucinations in Memory Systems of AgentsLLM agents and tool use → Medical and scientific domain agents
ROSBag MCP Server: Analyzing Robot Data with LLMs for Agentic Embodied AI ApplicationsLLM agents and tool use → Multimodal and embodied agents
SVG Decomposition for Enhancing Large Multimodal Models Visualization Comprehension: A Study with Floor PlansLLM reasoning → Spatial and embodied reasoning +1
ContextPilot: Fast Long-Context Inference via Context ReuseEfficient LLM inference and serving → Efficient reasoning and agent inference
Reinforcement Learning Using known InvariancesDeep reinforcement learning methods → RL for robot locomotion and manipulation
CareMedEval dataset: Evaluating Critical Appraisal and Reasoning in the Biomedical FieldLLM reasoning → Clinical and biomedical applications
Inter-Agent Trust Models: A Comparative Study of Brief, Claim, Proof, Stake, Reputation and Constraint in Agentic Web Protocol Design-A2A, AP2, ERC-8004, and BeyondLLM agents and tool use → Agent threats, guardrails and governance
Efficient Reasoning via Thought-Training and Thought-Free InferenceLLM reasoning → Test-time compute and efficient thinking
Towards Realistic Project-Level Code Generation via Multi-Agent Collaboration and Semantic Architecture ModelingLLM agents and tool use → Software engineering coding agents
LFC-DA: Logical Formula-Controlled Data Augmentation for Enhanced Logical ReasoningLLM reasoning → Reasoning benchmarks and evaluation
Silenced Biases: The Dark Side LLMs Learned to RefuseLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Decoupling Augmentation Bias in Prompt Learning for Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
A Modular, Data-Free Pipeline for Multi-Label Intention Recognition in Transportation Agentic AI ApplicationsLLM agents and tool use → General LLM agent frameworks and benchmarks
Benchmarking the Thinking Mode of Multimodal Large Language Models in Clinical TasksLLM reasoning → Clinical and biomedical applications +1
Diffusion-SDPO: Safeguarded Direct Preference Optimization for Diffusion ModelsDiffusion models → Diffusion and flow policies for robotics +1
Multi-Objective Adaptive Rate Limiting in Microservices Using Deep Reinforcement LearningDeep reinforcement learning methods → RL for energy, networks and operations
Diffusion Language Models are Super Data LearnersDiffusion models → Discrete and text diffusion
Unified Long Video Inpainting and Outpainting via Overlapping High-Order Co-DenoisingDiffusion models → 3D, point cloud and scene generation
Let the Bees Find the Weak Spots: A Path Planning Perspective on Multi-Turn Jailbreak Attacks against LLMsLLM safety and security → Jailbreaking
Comparing the Performance of LLMs in RAG-based Question-Answering: A Case Study in Computer Science LiteratureRetrieval-augmented generation → RAG pipeline design and evaluation
GMoPE:A Prompt-Expert Mixture Framework for Graph Foundation ModelsMixture-of-experts → MoE for time series and graphs
Death by a Thousand Prompts: Open Model Vulnerability AnalysisLLM safety and security → Safety alignment, refusal and guardrails
Incorporating Quality of Life in Climate Adaptation Planning via Reinforcement LearningDeep reinforcement learning methods → RL for energy, networks and operations
Transformer-Progressive Mamba Network for Lightweight Image Super-ResolutionState-space sequence models (S4, Mamba) → Vision Mamba backbones
Hybrid Fact-Checking that Integrates Knowledge Graphs, Large Language Models, and Search-Based Retrieval Agents Improves Interpretable Claim VerificationLLM agents and tool use → Web, search, and research agents
QG-CoC: Question-Guided Chain-of-Captions for Large Multimodal ModelsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Provable Separations between Memorization and Generalization in Diffusion ModelsDiffusion models → Diffusion foundations and sampling
Statistical Properties of Rectified FlowDiffusion models → Flow matching, time series and science
Periodic Skill DiscoveryDeep reinforcement learning methods → Deep RL algorithms and representations
BengaliMoralBench: A Benchmark for Auditing Moral Reasoning in Large Language Models within Bengali Language and CultureLLM reasoning → Reasoning benchmarks and evaluation
Toward Autonomous Engineering Design: A Knowledge-Guided Multi-Agent FrameworkLLM agents and tool use → Multimodal and embodied agents
Learning Natural and Robust Hexapod Locomotion over Complex Terrains via Motion Priors based on Deep Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Finetuning-Free Personalization of Text to Image Generation via HypernetworksDiffusion models → Image generation, text-to-image and editing
RefAgent: A Multi-agent LLM-based Framework for Automatic Software RefactoringLLM agents and tool use → Software engineering coding agents
MME-CC: A Challenging Multi-Modal Evaluation Benchmark of Cognitive CapacityLLM reasoning → Multimodal reasoning
From Measurement to Expertise: Empathetic Expert Adapters for Context-Based Empathy in Conversational AI AgentsLLM agents and tool use → Social simulation and multi-agent debate
DeepKnown-Guard: A Proprietary Model-Based Safety Response Framework for AI AgentsLLM agents and tool use → Agent threats, guardrails and governance
Optimal Boundary Control of Diffusion on Graphs via Linear ProgrammingDiffusion models → Flow matching, time series and science
Control Barrier Function for Aligning Large Language ModelsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks
Scaling Multi-Agent Environment Co-Design with Diffusion ModelsDiffusion models → Diffusion and flow policies for robotics
ALAS: Transactional and Dynamic Multi-Agent LLM PlanningLLM agents and tool use → Software engineering coding agents
SnapStream: Efficient Long Sequence Decoding on Dataflow AcceleratorsEfficient LLM inference and serving → KV cache optimization
EvalCards: A Framework for Standardized Evaluation Reporting
Learning Interestingness in Automated Mathematical Theory Formation
COMPASS: Context-Modulated PID Attention Steering System for Hallucination Mitigation
Extended Physics Informed Neural Network for Hyperbolic Two-Phase Flow in Porous Media
THD-BAR: Topology Hierarchical Derived Brain Autoregressive Modeling for EEG Generic Representations
Hierarchical Federated Graph Attention Networks for Scalable and Resilient UAV Collision Avoidance
Evaluating Modern Large Language Models on Low-Resource and Morphologically Rich Languages:A Cross-Lingual Benchmark Across Cantonese, Japanese, and Turkish
A Super-Learner with Large Language Models for Medical Emergency Advising
Assessing Identity Leakage in Talking Face Generation: Metrics and Evaluation Framework
Learning based Modelling of Throttleable Engine Dynamics for Lunar Landing Mission
MoE-GraphSAGE-Based Integrated Evaluation of Transient Rotor Angle and Voltage Stability in Power Systems
Case Study: Transformer-Based Solution for the Automatic Digitization of Gas Plants
Data-driven Feynman-Kac Discovery with Applications to Prediction and Data Generation
Exploring the Psychometric Validity of AI-Generated Student Responses: A Study on Virtual Personas' Learning Motivation
Large Language Models for Scientific Idea Generation: A Creativity-Centered Survey
From Prompts to Power: Measuring the Energy Footprint of LLM Inference
AutoHood3D: A Multi-Modal Benchmark for Automotive Hood Design and Fluid-Structure Interaction
FlowNet: Modeling Dynamic Spatio-Temporal Systems via Flow Propagation
Optimizing Predictive Maintenance in Intelligent Manufacturing: An Integrated FNO-DAE-GNN-PPO MDP Framework

All 290