Papers

New on arXiv

Tue 26 Aug 2025

357 papers

ProtSAE: Disentangling and Interpreting Protein Language Models via Semantically-Guided Sparse AutoencodersMechanistic interpretability → Interpreting biomedical and science models
Entropy-Guided Loop: Achieving Reasoning through Uncertainty-Aware GenerationLLM reasoning → Test-time compute and efficient thinking
ChipChat: Low-Latency Cascaded Conversational Agent in MLXLLM agents and tool use → General LLM agent frameworks and benchmarks
Test of Time: Rethinking Temporal Signal of Benchmark ContaminationLLM reasoning → Test-time compute and efficient thinking
Scaffold Diffusion: Sparse Multi-Category Voxel Structure Generation with Discrete DiffusionDiffusion models → 3D, point cloud and scene generation
QAgent: An LLM-based Multi-Agent System for Autonomous OpenQASM programmingLLM agents and tool use → Medical and scientific domain agents
ArgRAG: Explainable Retrieval Augmented Generation using Quantitative Bipolar ArgumentationRetrieval-augmented generation → RAG robustness and adaptive retrieval
Gentle Object Retraction in Dense Clutter Using Multimodal Force Sensing and Imitation LearningOffline RL and imitation learning from data → Robot manipulation imitation learning
Reliable Weak-to-Strong Monitoring of LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance +1
The Sample Complexity of Membership Inference and Privacy AuditingLLM safety and security → Membership inference attacks
Heterogeneous LLM Methods for Ontology Learning (Few-Shot Prompting, Ensemble Typing, and Attention-Based Taxonomies)Retrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
Fine-Tuning Vision-Language Models for Neutrino Event Analysis in High-Energy Physics ExperimentsLLM reasoning → Multimodal reasoning +1
HAP: Hybrid Adaptive Parallelism for Efficient Mixture-of-Experts InferenceMixture-of-experts → Efficient MoE training and inference systems
Grounding the Ungrounded: A Spectral-Graph Framework for Quantifying Hallucinations in Multimodal LLMsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
LongReasonArena: A Long Reasoning Benchmark for Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
Atrial Fibrillation Prediction Using a Lightweight Temporal Convolutional and Selective State Space ArchitectureState-space sequence models (S4, Mamba) → Physiological signal modeling
Reflective Agreement: Combining Self-Mixture of Agents with a Sequence Tagger for Robust Event ExtractionLLM agents and tool use → Agent memory systems
Context-Adaptive Synthesis and Compression for Enhanced Retrieval-Augmented Generation in Complex DomainsRetrieval-augmented generation → Generation and context integration
Re:Frame -- Retrieving Experience From Associative MemoryOffline RL and imitation learning from data → Offline RL algorithms +1
AT-CXR: Uncertainty-Aware Agentic Triage for Chest X-raysLLM agents and tool use → Medical and scientific domain agents
(DEMO) Deep Reinforcement Learning Based Resource Allocation in Distributed IoT SystemsDeep reinforcement learning methods → RL for energy, networks and operations
Model Context Protocols in Adaptive Transport Systems: A SurveyLLM agents and tool use → Tool learning and function calling
MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic ManipulationVision-language-action models → Reasoning and chain-of-thought in VLA
Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning
Understanding Tool-Integrated ReasoningLLM reasoning → Reasoning for agents, planning and code +1
The Ramon Llull's Thinking Machine for Automated IdeationLLM reasoning → Reasoning benchmarks and evaluation
All-in-One Slider for Attribute Manipulation in Diffusion ModelsDiffusion models → Image generation, text-to-image and editing
MATRIX: Multi-Agent simulaTion fRamework for safe Interactions and conteXtual clinical conversational evaluationLLM agents and tool use → Medical and scientific domain agents
RDDM: Practicing RAW Domain Diffusion Model for Real-world Image RestorationDiffusion models → Inverse problems and image restoration
QuadKAN: KAN-Enhanced Quadruped Motion Control via End-to-End Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
Active Query Selection for Crowd-Based Reinforcement LearningDeep reinforcement learning methods → RL training of language and agent models
Hybrid Deep Searcher: Scalable Parallel and Sequential Search ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning +2
Do LVLMs Know What They Know? A Systematic Study of Knowledge Boundary Perception in LVLMsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Composition and Alignment of Diffusion Models using Constrained LearningDiffusion models → Diffusion foundations and sampling
Reasoning LLMs in the Medical Domain: A Literature SurveyLLM reasoning → Clinical and biomedical applications
Trustworthy Agents for Electronic Health Records through Confidence EstimationLLM agents and tool use → Medical and scientific domain agents
Retrieval-Augmented Generation for Natural Language Art Provenance Searches in the Getty Provenance IndexRetrieval-augmented generation → Multimodal RAG
Federated Fine-Tuning of Sparsely-Activated Large Language Models on Resource-Constrained DevicesMixture-of-experts → Federated mixture-of-experts
HiPlan: Hierarchical Planning for LLM-Based Agents with Adaptive Global-Local GuidanceLLM agents and tool use → General LLM agent frameworks and benchmarks
Attackers Strike Back? Not Anymore -- An Ensemble of RL Defenders Awakens for APT DetectionDeep reinforcement learning methods → Adversarial robustness and RL in finance
Tackling Federated Unlearning as a Parameter Estimation ProblemLLM safety and security → Federated unlearning and poisoning
A Concurrent Modular Agent: Framework for Autonomous LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Investigating Advanced Reasoning of Large Language Models via Black-Box Environment InteractionLLM reasoning → Reasoning benchmarks and evaluation
MovieCORE: COgnitive REasoning in MoviesLLM reasoning → Multimodal reasoning +1
GitTaskBench: A Benchmark for Code Agents Solving Real-World Tasks Through Code Repository LeveragingLLM agents and tool use → Software engineering coding agents
Ask Me Again Differently: GRAS for Measuring Bias in Vision Language Models on Gender, Race, Age, and Skin ToneMultimodal LLMs / vision-language models → Social bias and fairness in VLMs
Interpretable by AI Mother Tongue: Native Symbolic Reasoning in Neural ModelsLLM reasoning → Reasoning benchmarks and evaluation
Enhancing Document VQA Models via Retrieval-Augmented GenerationMultimodal LLMs / vision-language models → Document understanding and multimodal RAG +1
SMoE: An Algorithm-System Co-Design for Pushing MoE to the Edge via Expert SubstitutionMixture-of-experts → Efficient MoE training and inference systems
Energy-Based Flow Matching for Generating 3D Molecular StructureDiffusion models → Flow matching, time series and science
Diverse And Private Synthetic Datasets Generation for RAG evaluation: A multi-agent frameworkRetrieval-augmented generation → Privacy and security of RAG +2
Interpretable Decision-Making for End-to-End Autonomous DrivingEnd-to-end autonomous driving → End-to-end system architecture
Toward Robust Medical Fairness: Debiased Dual-Modal Alignment via Text-Guided Attribute-Disentangled Prompt Learning for Vision-Language ModelsMultimodal LLMs / vision-language models → Medical vision-language models
Challenges for Generative AI in Legal ReasoningLLM reasoning → Reasoning for agents, planning and code
DRMD: Deep Reinforcement Learning for Malware Detection under Concept DriftDeep reinforcement learning methods → Adversarial robustness and RL in finance
On the Application of Diffusion Models for Simultaneous Denoising and DereverberationDiffusion models → Audio, speech and music
Arrows of Math Reasoning Data Synthesis for Large Language Models: Diversity, Complexity and CorrectnessLLM reasoning → Mathematical reasoning +1
Learning Acrobatic Flight from PreferencesDeep reinforcement learning methods → RL for robot navigation and UAV control
STARec: An Efficient Agent Framework for Recommender Systems via Autonomous Deliberate ReasoningLLM reasoning → LLM reasoning for recommendation +1
Hidden Tail: Adversarial Image Causing Stealthy Resource Consumption in Vision-Language ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
CausalMACE: Causality Empowered Multi-Agents in Minecraft Cooperative TasksLLM agents and tool use → Reinforcement learning for LLM agents
Harnessing Rule-Based Reinforcement Learning for Enhanced Grammatical Error CorrectionDeep reinforcement learning methods → RL training of language and agent models
ThinkDial: An Open Recipe for Controlling Reasoning Effort in Large Language ModelsLLM reasoning → Test-time compute and efficient thinking +1
Dynamic Collaboration of Multi-Language Models based on Minimal Complete Semantic UnitsLLM reasoning → Test-time compute and efficient thinking
Answering the Unanswerable Is to Err Knowingly: Analyzing and Mitigating Abstention Failures in Large Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
UltraMemV2: Memory Networks Scaling to 120B Parameters with Superior Long-Context LearningMixture-of-experts → MoE large language models
CrossHOI-Bench: A Unified Benchmark for HOI Evaluation across Vision-Language Models and HOI-Specific MethodsMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection
Chronological Passage Assembling in RAG framework for Temporal Question AnsweringRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
CAC-CoT: Connector-Aware Compact Chain-of-Thought for Efficient Reasoning Data Synthesis Across Dual-System Cognitive TasksLLM reasoning → Prompting and chain-of-thought reasoning +1
Stability and Generalization for Bellman ResidualsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +2
Rethinking Caching for LLM Serving Systems: Beyond Traditional HeuristicsLLM reasoning → Test-time compute and efficient thinking +1
Bias Mitigation Agent: Optimizing Source Selection for Fair and Balanced Knowledge RetrievalLLM agents and tool use → Web, search, and research agents
VistaWise: Building Cost-Effective Agent with Cross-Modal Knowledge Graph for MinecraftLLM agents and tool use → Reinforcement learning for LLM agents
AppAgent-Pro: A Proactive GUI Agent System for Multidomain Information Integration and User AssistanceLLM agents and tool use → GUI and mobile agents
Utilizing Training Data to Improve LLM Reasoning for Tabular UnderstandingLLM reasoning → Knowledge- and tool-augmented reasoning
Requirements Development and Formalization for Reliable Code Generation: A Multi-Agent VisionLLM agents and tool use → Software engineering coding agents
Tailored Teaching with Balanced Difficulty: Elevating Reasoning in Multimodal Chain-of-Thought via Prompt CurriculumLLM reasoning → Prompting and chain-of-thought reasoning
Optimal Sparsity of Mixture-of-Experts Language Models for Reasoning TasksLLM reasoning → Test-time compute and efficient thinking +1
Auditing Approximate Machine Unlearning for Differentially Private ModelsLLM safety and security → Machine unlearning
MUA-RL: Multi-turn User-interacting Agent Reinforcement Learning for agentic tool useLLM agents and tool use → Reinforcement learning for LLM agents
Membership Inference Attacks on In-Context Examples in LLM-based Recommender SystemsLLM safety and security → Membership inference attacks
FFT-MoE: Efficient Federated Fine-Tuning for Foundation Models via Large-scale Sparse MoE under Heterogeneous EdgeMixture-of-experts → Federated mixture-of-experts
UniC-RAG: Universal Knowledge Corruption Attacks to Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG
PRISM: Robust VLM Alignment with Principled Reasoning for Integrated Safety in MultimodalityLLM reasoning → Reasoning for agents, planning and code +2
Thinking Before You Speak: A Proactive Test-time Scaling ApproachLLM reasoning → Reasoning benchmarks and evaluation
RLMR: Reinforcement Learning with Mixed Rewards for Creative WritingLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment +1
STRATA-TS: Selective Knowledge Transfer for Urban Time Series Forecasting with Retrieval-Guided ReasoningLLM reasoning → Time series reasoning
Decouple, Reorganize, and Fuse: A Multimodal Framework for Cancer Survival PredictionMixture-of-experts → Medical and biomedical MoE
Bias-Adjusted LLM Agents for Human-Like Decision-Making via Behavioral EconomicsLLM agents and tool use → Social simulation and multi-agent debate
What do language models model? Transformers, automata, and the format of thoughtLLM reasoning → Reasoning benchmarks and evaluation
SemLayoutDiff: Semantic Layout Generation with Diffusion Model for Indoor Scene SynthesisDiffusion models → 3D, point cloud and scene generation
History Rhymes: Accelerating LLM Reinforcement Learning with RhymeRLLLM reasoning → Reinforcement learning for reasoning
DrugReasoner: Interpretable Drug Approval Prediction with a Reasoning-augmented Language ModelLLM reasoning → Clinical and biomedical applications
Strata: Hierarchical Context Caching for Long Context Language Model ServingEfficient LLM inference and serving → KV cache optimization
The Mind's Eye: A Multi-Faceted Reward Framework for Guiding Visual Metaphor GenerationLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
Development and Evaluation of an AI-Driven Telemedicine System for Prenatal Healthcare
A Masked Representation Learning to Model Cardiac Functions Using Multiple Physiological Signals
Topic Identification in LLM Input-Output Pairs through the Lens of Information Bottleneck
Real-Time Detection of Hallucinated Entities in Long-Form Generation
Latent Variable Modeling in Multi-Agent Reinforcement Learning via Expectation-Maximization for UAV-Based Wildlife Protection

All 357