Papers

New on arXiv

Thu 6 Nov 2025

337 papers

Ask WhAI:Probing Belief Formation in Role-Primed LLM AgentsLLM agents and tool use → Medical and scientific domain agents
Guarding the Meaning: Self-Supervised Training for Semantic Robustness in Guard ModelsLLM safety and security → Safety alignment, refusal and guardrails
Reasoning on Time-Series for Financial Technical AnalysisLLM reasoning → Reasoning for agents, planning and code
GRIP: In-Parameter Graph Reasoning through Fine-Tuning Large Language ModelsLLM reasoning → Knowledge- and tool-augmented reasoning
Personalized Image Editing in Text-to-Image Diffusion Models via Collaborative Direct Preference OptimizationDiffusion models → Image generation, text-to-image and editing +1
Walking the Schr"odinger Bridge: A Direct Trajectory for Text-to-3D GenerationDiffusion models → 3D, point cloud and scene generation
FiCABU: A Fisher-Based, Context-Adaptive Machine Unlearning Processor for Edge AILLM safety and security → Machine unlearning
DMA: Online RAG Alignment with Human FeedbackRetrieval-augmented generation → RAG pipeline design and evaluation +1
FoodRL: A Reinforcement Learning Ensembling Framework For In-Kind Food Donation ForecastingDeep reinforcement learning methods → RL for energy, networks and operations
Quantum Boltzmann Machines for Sample-Efficient Reinforcement LearningDeep reinforcement learning methods → Quantum reinforcement learning
Test-Time Adaptation for LLM Agents via Environment InteractionLLM agents and tool use → General LLM agent frameworks and benchmarks
Sublinear iterations can suffice even for DDPMsDiffusion models → Diffusion foundations and sampling
Prompt-Based Safety Guidance Is Ineffective for Unlearned Text-to-Image Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects +1
Agentic Refactoring: An Empirical Study of AI Coding AgentsLLM agents and tool use → Software engineering coding agents
Multimodal Diffusion Forcing for Forceful ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
PuzzleMoE: Efficient Compression of Large Mixture-of-Experts Models via Sparse Expert Merging and Bit-packed inferenceMixture-of-experts → MoE large language models
Explore Data Left Behind in Reinforcement Learning for Reasoning Language ModelsLLM reasoning → Reinforcement learning for reasoning
DuetServe: Harmonizing Prefill and Decode for LLM Serving via Adaptive GPU MultiplexingEfficient LLM inference and serving → Serving systems and request scheduling
CPO: Condition Preference Optimization for Controllable Image GenerationLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
IndicVisionBench: Benchmarking Cultural and Multilingual Understanding in VLMsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Learning to reason about rare diseases through retrieval-augmented agentsLLM reasoning → Clinical and biomedical applications +2
P-MIA: A Profiled-Based Membership Inference Attack on Cognitive Diagnosis ModelsLLM safety and security → Membership inference attacks
X-Diffusion: Training Diffusion Policies on Cross-Embodiment Human DemonstrationsDiffusion models → Diffusion and flow policies for robotics
Cambrian-S: Towards Spatial Supersensing in VideoWorld models → Video-generation world models and simulators
SIMS-V: Simulated Instruction-Tuning for Spatial Video UnderstandingLLM reasoning → Spatial and embodied reasoning +1
VeriCoT: Neuro-symbolic Chain-of-Thought Validation via Logical Consistency ChecksLLM reasoning → Reasoning benchmarks and evaluation
Benchmark Designers Should "Train on the Test Set" to Expose Exploitable Non-Visual ShortcutsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Logit-Entropy Adaptive Stopping Heuristic for Efficient Chain-of-Thought ReasoningLLM reasoning → Test-time compute and efficient thinking
Optimal Inference Schedules for Masked Diffusion ModelsDiffusion models → Discrete and text diffusion
DR. WELL: Dynamic Reasoning and Learning with Symbolic World Model for Embodied LLM-Based Multi-Agent CollaborationWorld models → World models for language agents +1
Efficient probabilistic surrogate modeling techniques for partially-observed large-scale dynamical systemsDiffusion models → Flow matching, time series and science
Addressing divergent representations from causal interventions on neural networksMechanistic interpretability → Circuits and mechanistic analysis
Environment Agnostic Goal-Conditioning, A Study of Reward-Free Autonomous LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Thinking with Video: Video Generation as a Promising Multimodal Reasoning ParadigmLLM reasoning → Multimodal reasoning +1
BanglaMedQA and BanglaMMedBench: Evaluating Retrieval-Augmented Generation Strategies for Bangla Biomedical Question AnsweringRetrieval-augmented generation → Medical and biomedical RAG +1
Evo-1: Lightweight Vision-Language-Action Model with Preserved Semantic AlignmentVision-language-action models → Generalist VLA models and benchmarks
Are language models aware of the road not taken? Token-level uncertainty and hidden state dynamicsLLM reasoning → Prompting and chain-of-thought reasoning
End-to-End Reinforcement Learning of Koopman Models for eNMPC of an Air Separation UnitDeep reinforcement learning methods → RL for autonomous driving and traffic
Large Language Models for Cyber SecurityLLM safety and security → Prompt injection and agent security
RAGalyst: Automated Human-Aligned Agentic Evaluation for Domain-Specific RAGRetrieval-augmented generation → Agentic RAG, agents and memory +1
RUST-BENCH: Benchmarking LLM Reasoning on Unstructured Text within Structured TablesLLM reasoning → Knowledge- and tool-augmented reasoning
Promoting Sustainable Web Agents: Benchmarking and Estimating Energy Consumption through Empirical and Theoretical AnalysisLLM agents and tool use → Web, search, and research agents
ThaiOCRBench: A Task-Diverse Benchmark for Vision-Language Understanding in ThaiMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Generate, Evaluate, Iterate: Synthetic Data for Human-in-the-Loop Refinement of LLM JudgesSynthetic data and self-improvement for LLMs → LLM synthetic data generation
Beyond Shortest Path: Agentic Vehicular Routing with Semantic ContextLLM agents and tool use → Multimodal and embodied agents
V-Thinker: Interactive Thinking with ImagesLLM reasoning → Multimodal reasoning +2
Fitting Reinforcement Learning Model to Behavioral Data under BanditsDeep reinforcement learning methods → RL theory and sample complexity
CoRPO: Adding a Correctness Bias to GRPO Improves GeneralizationLLM reasoning → Reinforcement learning for reasoning
If I Could Turn Back Time: Temporal Reframing as a Historical Reasoning Task for LLMsLLM reasoning → Knowledge- and tool-augmented reasoning
Speed at the Cost of Quality: How Cursor AI Increases Short-Term Velocity and Long-Term Complexity in Open-Source ProjectsLLM agents and tool use → Software engineering coding agents
Temporal Action Selection for Action ChunkingOffline RL and imitation learning from data → Robot manipulation imitation learning
Post-Training LLMs as Better Decision-Making Agents: A Regret-Minimization ApproachLLM agents and tool use → Reinforcement learning for LLM agents
MusRec: Zero-Shot Text-to-Music Editing via Rectified Flow and Diffusion TransformersDiffusion models → Audio, speech and music
Towards Aligning Multimodal LLMs with Human Experts: A Focus on Parent-Child InteractionMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
GraSP-VLA: Graph-based Symbolic Action Representation for Long-Horizon Planning with VLA PoliciesVision-language-action models → Generalist VLA models and benchmarks
Monitor-Generate-Verify (MGV): Formalising Metacognitive Theory for Language Model ReasoningLLM reasoning → Test-time compute and efficient thinking
RISE-T2V: Rephrasing and Injecting Semantics with LLM for Expansive Text-to-Video GenerationDiffusion models → Video generation
AdversariaLLM: A Unified and Modular Toolbox for LLM Robustness ResearchLLM safety and security → Jailbreaking
DeepPAAC: A New Deep Galerkin Method for Principal-Agent ProblemsDeep reinforcement learning methods → RL theory and sample complexity
GUI-360^circ: A Comprehensive Dataset and Benchmark for Computer-Using AgentsLLM agents and tool use → GUI and mobile agents
Efficient Reinforcement Learning from Human Feedback via Bayesian Preference InferenceLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
RLoop: An Self-Improving Framework for Reinforcement Learning with Iterative Policy InitializationLLM reasoning → Reinforcement learning for reasoning +1
SSPO: Subsentence-level Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
Can Context Bridge the Reality Gap? Sim-to-Real Transfer of Context-Aware PoliciesDeep reinforcement learning methods → Safe and constrained RL
REMIND: Input Loss Landscapes Reveal Residual Memorization in Post-Unlearning LLMsLLM safety and security → Machine unlearning
Black-Box Guardrail Reverse-engineering AttackLLM safety and security → Safety alignment, refusal and guardrails
Learning to Land Anywhere: Transferable Generative Models for Aircraft TrajectoriesDiffusion models → Diffusion and flow policies for robotics
BAPPA: Benchmarking Agents, Plans, and Pipelines for Automated Text-to-SQL GenerationLLM agents and tool use → Web, search, and research agents
Exchange Policy Optimization Algorithm for Semi-Infinite Safe Reinforcement LearningDeep reinforcement learning methods → Safe and constrained RL
Learning from Online Videos at Inference Time for Computer-Use AgentsLLM agents and tool use → GUI and mobile agents
BFM-Zero: A Promptable Behavioral Foundation Model for Humanoid Control Using Unsupervised Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Text to Sketch Generation with Multi-StylesMultimodal LLMs / vision-language models → Image editing and quality assessment
RIDE: Difficulty Evolving Perturbation with Item Response Theory for Mathematical ReasoningLLM reasoning → Mathematical reasoning
Tortoise and Hare Guidance: Accelerating Diffusion Model Inference with Multirate IntegrationDiffusion models → Diffusion foundations and sampling
Batch Prompting Suppresses Overthinking Reasoning Under Constraint: How Batch Prompting Suppresses Overthinking in Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
KGFR: A Foundation Retriever for Generalized Knowledge Graph Question AnsweringLLM reasoning → Knowledge- and tool-augmented reasoning
Caption Injection for Optimization in Generative Search EngineRetrieval-augmented generation → Multimodal RAG
Agentmandering: A Game-Theoretic Framework for Fair Redistricting via Large Language Model AgentsLLM agents and tool use → Social simulation and multi-agent debate
Plan of Knowledge: Retrieval-Augmented Large Language Models for Temporal Knowledge Graph Question AnsweringRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
Benchmarking and Studying the LLM-based Agent System in End-to-End Software DevelopmentLLM agents and tool use → Software engineering coding agents
Necessary and Sufficient Conditions for the Optimization-Based Concurrent Execution of Learned Robotic TasksDeep reinforcement learning methods → RL theory and sample complexity
Interpreting Multi-Attribute Confounding through Numerical Attributes in Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
PICNIC: Silicon Photonic Interconnected Chiplets with Computational Network and In-memory Computing for LLM Inference AccelerationEfficient LLM inference and serving → Serving systems and request scheduling
Detecting Silent Failures in Multi-Agentic AI TrajectoriesLLM agents and tool use → Software engineering coding agents
Abductive Inference in Retrieval-Augmented Language Models: Generating and Validating Missing PremisesRetrieval-augmented generation → RAG robustness and adaptive retrieval
VulInstruct: Teaching LLMs Root-Cause Reasoning for Vulnerability Detection via Security SpecificationsLLM reasoning → Reasoning for agents, planning and code
GNN-MoE: Context-Aware Patch Routing using GNNs for Parameter-Efficient Domain GeneralizationMixture-of-experts → Vision mixture-of-experts
PhysCorr: Dual-Reward DPO for Physics-Constrained Text-to-Video Generation with Automated Preference SelectionLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
ArchPilot: A Proxy-Guided Multi-Agent Approach for Machine Learning EngineeringLLM agents and tool use → Web, search, and research agents
PrivacyCD: Hierarchical Unlearning for Protecting Student Privacy in Cognitive DiagnosisLLM safety and security → Machine unlearning
Multi-Agent Collaborative Framework For Math Problem GenerationLLM agents and tool use → Educational agents and math problem solving
Direct Semantic Communication Between Large Language Models via Vector TranslationLLM agents and tool use → General LLM agent frameworks and benchmarks
RLHF: A comprehensive Survey for Cultural, Multimodal and Low Latency Alignment MethodsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
PEFA-AI: Advancing Open-source LLMs for RTL generation using Progressive Error Feedback Agentic-AILLM agents and tool use → Software engineering coding agents
SynQuE: Estimating Synthetic Dataset Quality Without AnnotationsSynthetic data and self-improvement for LLMs → LLM synthetic data generation
Collaborative Agents for Automated Program Repair in RubyLLM agents and tool use → Software engineering coding agents
A Survey on Improving Human Robot Collaboration through Vision-and-Language Navigation
HiFiNet: Hierarchical Fault Identification in Wireless Sensor Networks via Edge-Based Classification and Graph Aggregation
Quantifying the Role of OpenFold Components in Protein Structure Prediction
The Impact of Prosodic Segmentation on Speech Synthesis of Spontaneous Speech

All 337