Papers

New on arXiv

Fri 5 Dec 2025

306 papers

Graph Your Way to Inspiration: Integrating Co-Author Graphs with Retrieval-Augmented Generation for Large Language Model Based Scientific Idea GenerationRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
The Forgotten Shield: Safety Grafting in Parameter-Space for Medical MLLMsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
ReCollab: Retrieval-Augmented LLMs for Cooperative Ad-hoc Teammate ModelingWorld models → World models for language agents +1
Deep Reinforcement Learning Optimization for Uncertain Nonlinear Systems via Event-Triggered Robust Adaptive Dynamic ProgrammingDeep reinforcement learning methods → RL theory and sample complexity
The Red Queen's Trap: Limits of Deep Evolution in High-Frequency TradingDeep reinforcement learning methods → Adversarial robustness and RL in finance
FedSight AI: Multi-Agent System Architecture for Federal Funds Target Rate PredictionLLM agents and tool use → Financial trading and forecasting agents
Mitigating Catastrophic Forgetting in Mathematical Reasoning Finetuning through Mixed TrainingLLM reasoning → Mathematical reasoning
Scaling and Transferability of Annealing Strategies in Large Language Model TrainingMixture-of-experts → MoE large language models
Adjudicator: Correcting Noisy Labels with a KG-Informed Council of LLM AgentsLLM agents and tool use → Web, search, and research agents
Safe2Harm: Semantic Isomorphism Attacks for Jailbreaking Large Language ModelsLLM safety and security → Jailbreaking
Explainable Adversarial-Robust Vision-Language-Action Model for Robotic ManipulationVision-language-action models → Safety and adversarial attacks
Consist-Retinex: One-Step Noise-Emphasized Consistency Training Accelerates High-Quality Retinex EnhancementDiffusion models → Inverse problems and image restoration
Training Multi-Image Vision Agents via End2End Reinforcement LearningLLM agents and tool use → Multimodal and embodied agents +1
What Happens When: Learning Temporal Orders of Events in VideosMultimodal LLMs / vision-language models → Video understanding with MLLMs
MARINE: Theoretical Optimization and Design for Multi-Agent Recursive IN-context EnhancementLLM reasoning → Reasoning for agents, planning and code +1
Average-reward reinforcement learning in semi-Markov decision processes via relative value iterationDeep reinforcement learning methods → RL theory and sample complexity
Quantifying Memory Use in Reinforcement Learning with Temporal RangeDeep reinforcement learning methods → Deep RL algorithms and representations
K2-V2: A 360-Open, Reasoning-Enhanced LLMLLM reasoning → Reasoning benchmarks and evaluation
How Should We Evaluate Data Deletion in Graph-Based ANN Indexes?Retrieval-augmented generation → Vector search and ANN indexes
ARCANE: A Multi-Agent Framework for Interpretable and Configurable AlignmentLLM agents and tool use → General LLM agent frameworks and benchmarks +1
PMA-Diffusion: A Physics-guided Mask-Aware Diffusion Framework for TSE from Sparse ObservationsDiffusion models → Diffusion and flow policies for robotics
WAM-Flow: Parallel Coarse-to-Fine Motion Planning via Discrete Flow Matching for Autonomous DrivingDiffusion models → Diffusion and flow policies for robotics +1
JaxWildfire: A GPU-Accelerated Wildfire Simulator for Reinforcement LearningDeep reinforcement learning methods → RL for energy, networks and operations
Empathy by Design: Aligning Large Language Models for Healthcare DialogueLLM reasoning → Clinical and biomedical applications +1
BeLLA: End-to-End Birds Eye View Large Language Assistant for Autonomous DrivingMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection +1
Reinforcement Learning Integrated Agentic RAG for Software Test Cases AuthoringLLM agents and tool use → Software engineering coding agents
Enhancing Retrieval-Augmented Generation with Entity Linking for Educational PlatformsRetrieval-augmented generation → RAG pipeline design and evaluation
EditThinker: Unlocking Iterative Reasoning for Any Image EditorLLM reasoning → Multimodal reasoning
Training-Time Action Conditioning for Efficient Real-Time ChunkingVision-language-action models → Efficient VLA inference and adaptation
Whatever Remains Must Be True: Filtering Drives Reasoning in LLMs, Shaping DiversityLLM reasoning → Reinforcement learning for reasoning
M4-RAG: A Massive-Scale Multilingual Multi-Cultural Multimodal RAGMultimodal LLMs / vision-language models → Document understanding and multimodal RAG +1
SIMPACT: Simulation-Enabled Action Planning using Vision-Language ModelsLLM reasoning → Spatial and embodied reasoning +1
SymPyBench: A Dynamic Benchmark for Scientific Reasoning with Executable Python CodeLLM reasoning → Spatial and embodied reasoning
Correspondence-Oriented Imitation Learning: Flexible Visuomotor Control with 3D ConditioningOffline RL and imitation learning from data → Robot manipulation imitation learning
Trusted AI Agents in the CloudLLM agents and tool use → Agent threats, guardrails and governance
TRACE: A Framework for Analyzing and Enhancing Stepwise Reasoning in Vision-Language ModelsLLM reasoning → Multimodal reasoning +1
PRiSM: An Agentic Multimodal Benchmark for Scientific Reasoning via Python-Grounded EvaluationLLM reasoning → Spatial and embodied reasoning +2
World Models That Know When They Don't Know - Controllable Video Generation with Calibrated UncertaintyWorld models → Embodied and robot world models
KQ-SVD: Compressing the KV Cache with Provable Guarantees on Attention FidelityEfficient LLM inference and serving → KV cache optimization
Bootstrapping Fuzzers for Compilers of Low-Resource Language Dialects Using Language ModelsSynthetic data and self-improvement for LLMs → Code and instruction data synthesis
Optimizing Medical Question-Answering Systems: A Comparative Study of Fine-Tuned and Zero-Shot Large Language Models with RAG FrameworkRetrieval-augmented generation → Medical and biomedical RAG
VRSA: Jailbreaking Multimodal Large Language Models through Visual Reasoning Sequential AttackLLM reasoning → Reasoning for agents, planning and code +2
NEAT: Neighborhood-Guided, Efficient, Autoregressive Set Transformer for 3D Molecular GenerationDiffusion models → Flow matching, time series and science
Toward Efficient and Robust Behavior Models for Multi-Agent Driving SimulationEnd-to-end autonomous driving → Imitation-learning end-to-end planners
Probing the effectiveness of World Models for Spatial Reasoning through Test-time ScalingWorld models → World models for language agents +2
Mechanistic Interpretability of Antibody Language Models Using SAEsMechanistic interpretability → Interpreting biomedical and science models
Fast and Robust Diffusion Posterior Sampling for MR Image Reconstruction Using the Preconditioned Unadjusted Langevin AlgorithmDiffusion models → Medical imaging and segmentation
Active Video Perception: Iterative Evidence Seeking for Agentic Long Video UnderstandingLLM agents and tool use → Multimodal and embodied agents
Evolutionary System 2 Reasoning: An Empirical ProofLLM reasoning → Reasoning benchmarks and evaluation
USV: Unified Sparsification for Accelerating Video Diffusion ModelsDiffusion models → Video generation
A Fast Anti-Jamming Cognitive Radar Deployment Algorithm Based on Reinforcement LearningDeep reinforcement learning methods → RL for robot navigation and UAV control
Capturing Classic Authorial Style in Long-Form Story Generation with GRPO Fine-TuningLLM reasoning → Reasoning benchmarks and evaluation +1
HQ-DM: Single Hadamard Transformation-Based Quantization-Aware Training for Low-Bit Diffusion ModelsDiffusion models → Diffusion foundations and sampling
ARGUS: Defending Against Multimodal Indirect Prompt Injection via Steering Instruction-Following BehaviorMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
HiMoE-VLA: Hierarchical Mixture-of-Experts for Generalist Vision-Language-Action PoliciesVision-language-action models → Generalist VLA models and benchmarks +1
LA-RL: Language Action-guided Reinforcement Learning with Safety Guarantees for Autonomous Highway DrivingEnd-to-end autonomous driving → Large language models for driving +1
InverseCrafter: Efficient Video ReCapture as a Latent Domain Inverse ProblemDiffusion models → 3D, point cloud and scene generation
Interleaved Latent Visual Reasoning with Selective Perceptual ModelingLLM reasoning → Multimodal reasoning +1
Multilingual Medical Reasoning for Question Answering with Large Language ModelsLLM reasoning → Clinical and biomedical applications
Beyond Data Filtering: Knowledge Localization for Capability Removal in LLMsLLM safety and security → Machine unlearning
A Greek Government Decisions Dataset for Public-Sector Analysis and InsightRetrieval-augmented generation → Legal and financial RAG
Entropy Ratio Clipping as a Soft Global Constraint for Stable Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
Structured Reasoning with Tree-of-Thoughts for Bengali Math Word ProblemsLLM reasoning → Mathematical reasoning
CureAgent: A Training-Free Executor-Analyst Framework for Clinical ReasoningLLM reasoning → Clinical and biomedical applications
MedDIFT: Multi-Scale Diffusion-Based Correspondence in 3D Medical ImagingDiffusion models → Medical imaging and segmentation
Conscious Gaze: Adaptive Attention Mechanisms for Hallucination Mitigation in Vision-Language ModelsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
RoBoN: Routed Online Best-of-n for Test-Time Scaling with Multiple LLMsLLM reasoning → Test-time compute and efficient thinking
A Unified Theory of Sparse Dictionary Learning in Mechanistic Interpretability: Piecewise Biconvexity and Spurious MinimaMechanistic interpretability → Sparse autoencoder training and evaluation
MIND: Multi-rationale INtegrated Discriminative Reasoning Framework for Multi-modal Large ModelsLLM reasoning → Multimodal reasoning +1
Matching Ranks Over Probability Yields Truly Deep Safety AlignmentLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Know-Show: Benchmarking Video-Language Models on Spatio-Temporal Grounded ReasoningLLM reasoning → Multimodal reasoning
GRASP: Graph Reasoning Agents for Systems Pharmacology with Human-in-the-LoopLLM reasoning → Clinical and biomedical applications +1
SEA-SafeguardBench: Evaluating AI Safety in SEA Languages and CulturesLLM safety and security → Safety alignment, refusal and guardrails
TeleAI-Safety: A comprehensive LLM jailbreaking benchmark towards attacks, defenses, and evaluationsLLM safety and security → Jailbreaking
Everything is Context: Agentic File System Abstraction for Context EngineeringLLM agents and tool use → Software engineering coding agents
Dynamic Alignment for Collective Agency: Toward a Scalable Self-Improving Framework for Open-Ended LLM AlignmentLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment +1
The Seeds of Scheming: Weakness of Will in the Building Blocks of Agentic SystemsLLM agents and tool use → General LLM agent frameworks and benchmarks
ArtistMus: A Globally Diverse, Artist-Centric Benchmark for Retrieval-Augmented Music Question AnsweringRetrieval-augmented generation → RAG pipeline design and evaluation
ParaUni: Enhance Generation in Unified Multimodal Model with Reinforcement-driven Hierarchical Parallel Information InteractionMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning +1
TS-HINT: Enhancing Semiconductor Time Series Regression Using Attention Hints From Large Language Model ReasoningLLM reasoning → Time series reasoning
A Systematic Framework for Enterprise Knowledge Retrieval: Leveraging LLM-Generated Metadata to Enhance RAG SystemsRetrieval-augmented generation → RAG pipeline design and evaluation
Delving into Latent Spectral Biasing of Video VAEs for Superior DiffusabilityDiffusion models → Diffusion foundations and sampling
LoC-Path: Learning to Compress for Pathology Multimodal Large Language ModelsMultimodal LLMs / vision-language models → Medical vision-language models
China Regional 3km Downscaling Based on Residual Corrective Diffusion ModelDiffusion models → Flow matching, time series and science
Please Don't Kill My Vibe: Empowering Agents with Data Flow ControlLLM agents and tool use → Agent threats, guardrails and governance
ChipMind: Retrieval-Augmented Reasoning for Long-Context Circuit Design SpecificationsLLM reasoning → Reasoning for agents, planning and code +1
MCP-AI: Protocol-Driven Intelligence Framework for Autonomous Reasoning in HealthcareLLM reasoning → Clinical and biomedical applications
FieldSeer I: Physics-Guided World Models for Long-Horizon Electromagnetic Dynamics under Partial ObservabilityWorld models → Latent predictive world models
When Forgetting Builds Reliability: LLM Unlearning for Reliable Hardware Code GenerationLLM safety and security → Machine unlearning
Taxonomy-Adaptive Moderation Model with Robust Guardrails for Large Language ModelsLLM safety and security → Safety alignment, refusal and guardrails
State-Conditional Adversarial Learning: An Off-Policy Visual Domain Transfer Method for End-to-End Imitation LearningOffline RL and imitation learning from data → Imitation learning methods
LYNX: Learning Dynamic Exits for Confidence-Controlled ReasoningLLM reasoning → Test-time compute and efficient thinking +1
FronTalk: Benchmarking Front-End Development as Conversational Code Generation with Multi-Modal Feedback
TeleTables: A Benchmark for Large Language Models in Telecom Table Interpretation
Collective Narrative Grounding: Community-Coordinated Data Contributions to Improve Local AI Systems
Attribute-Aware Controlled Product Generation with LLMs for E-commerce
A Context-Free Smart Grid Model Using Complex System Approach
TinyMyo: a Tiny Foundation Model for Flexible EMG Signal Processing at the Edge
Tourists Profiling by Interest Analysis
Modular connectivity in neural networks emerges from Poisson noise-motivated regularisation, and promotes robustness and compositional generalisation

All 306