Papers

New on arXiv

Sun 28 Jun 2026

253 papers

SHAPE of Chain-of-Thought in Math ReasoningLLM reasoning → Mathematical reasoning
A Survey on Foundations and Frontiers of Multimodal Agentic Frameworks: Techniques and ApplicationsLLM agents and tool use → Multimodal and embodied agents
From Naive RAG to Deep Agentic Retrieval: An Evolving Context Engineering Pipeline for Regulatory ComplianceRetrieval-augmented generation → Agentic RAG, agents and memory +1
Autonomous mechanistic discovery of colorectal cancer vulnerabilities via multi-scale AI swarmsLLM reasoning → Clinical and biomedical applications +1
AdaSurvMamba: Dynamic Fusion and Semantic Scanning for Multimodal Survival AnalysisState-space sequence models (S4, Mamba) → Mamba for medical image analysis
From Outcomes to Actions: Leveraging Hindsight for Long-Horizon Language Agent TrainingLLM agents and tool use → Reinforcement learning for LLM agents
Attention Dynamics in Diffusion Models: A Visual Analytics Framework for Human-AI CollaborationDiffusion models → Image generation, text-to-image and editing
Stop Hand-Holding Your Coding Agent: Engineering the Loops that Replace Step-by-Step PromptingLLM agents and tool use → Software engineering coding agents
Unsupervised Thermodynamics of Molecular Diffusion Models: Action-Operator Semantics and Auditable Free-Energy ReadoutDiffusion models → Flow matching, time series and science
Position: Vision-Language-Action Models Cannot Be Verified to Perform Physical ReasoningLLM reasoning → Spatial and embodied reasoning +2
Budgeted Act-or-Defer Multi-Agent LLM Deliberation with Local Reliability BoundsLLM reasoning → Reasoning for agents, planning and code +1
As We May SearchRetrieval-augmented generation → Privacy and security of RAG
Resolution Thresholds in VLM Detection of Harmful ASCII Art Across Construction Modes and LanguagesMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning +1
Hybrid Retriever Evolution for Multimodal Document Reasoning AgentsLLM reasoning → Knowledge- and tool-augmented reasoning
Fuzzing Large Language Models to Elicit Hidden BehavioursLLM safety and security → Backdoor and data poisoning attacks
Metadata, Structure, or Strategy? A Decomposition of RAG Context EnrichmentRetrieval-augmented generation → RAG pipeline design and evaluation
Fast Wireless Foundation Models with Early-ExitsEfficient LLM inference and serving → Early exit and edge inference
Two-Stage Prompt Optimization for Few-Shot Relation Extraction: From Reasoning-Guided Search to Gradient-Guided RefinementLLM reasoning → Prompting and chain-of-thought reasoning
Bidirectional Autoregressive Latent Diffusion for Forward and Inverse MagnetohydrodynamicsDiffusion models → Flow matching, time series and science
Does Role Specialization Matter for Explanation Faithfulness in Mixture-of-Experts?Mixture-of-experts → Expert routing strategies
An Empirical Evaluation of Prompt Injection Vulnerabilities in Large Language Models Across Multilingual and Obfuscated Attack ScenariosLLM safety and security → Prompt injection and agent security
SonoCLIP: Mask-Guided Region-Aware Vision-Language Pretraining for Fetal Ultrasound AnalysisMultimodal LLMs / vision-language models → Medical vision-language models
The Joint Effect of Quantization and Sampling Temperature on LLM Safety Alignment: A Factorial AnalysisLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
MAM-AI: An On-Device Medical Retrieval-Augmented Generation System for Nurses and Midwives in ZanzibarRetrieval-augmented generation → Medical and biomedical RAG
ScAle: Attention Head Scaling as a Minimal Adapter for Spatial Reasoning in Vision Language ModelsLLM reasoning → Spatial and embodied reasoning +1
TF-MoE: Time-Frequency Mixture-of-Experts for Efficient Speech SeparationMixture-of-experts → Multimodal MoE models
Reliability-Prioritized Fine-Grained Generation in Multimodal LargeMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
Coverage-Driven KV Cache Eviction for Efficient and Improved Inference of LLMEfficient LLM inference and serving → KV cache optimization
OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World TasksLLM agents and tool use → GUI and mobile agents
When Does Activation Steering Change What a Model Computes From?Mechanistic interpretability → Activation steering methods
CORE: Common Outcome Regularities from Action-Free Visual Demonstrations for Robot ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
Reinforcement Learning in Super Mario Bros: Curriculum, Pedagogy, and Optimal Level Design in World 1-1Deep reinforcement learning methods → RL for games and combinatorial optimization
UCOB: Learning to Utilize and Evolve Agentic Skills via Credit-Aware On-Policy Bidirectional Self-DistillationLLM agents and tool use → Agent skill learning and evolution +1
Learning Transferable Dynamics Priors from Action to World ModelingWorld models → Embodied and robot world models
Cognitive World Model for Progressive BDI/E Trajectory Evaluation of Conversational AgentsWorld models → World models for language agents +1
To Reason or to Fabricate: Reasoning Without Shortcuts via Hint-Anchored Pairwise AggregationLLM reasoning → Reinforcement learning for reasoning
CRAFT: Counterfactual Credit Assignment from Free Sibling Rollouts for Self-Distilled Agentic Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents +1
Agent-Computer Observation Interfaces Enable Dynamic Computer UseLLM agents and tool use → GUI and mobile agents
mamabench and mamaretrieval: Benchmarks for Evaluating Medical Retrieval-Augmented Generation in Maternal, Neonatal, and Reproductive HealthRetrieval-augmented generation → Medical and biomedical RAG
Prototype Latent World Model Replay for Class-Incremental LearningWorld models → Latent predictive world models
Rank-Aware Hyperbolic Alignment for Vision-Language Dataset DistillationMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
MIRROR: Aligning Semantic Relations from Language to Image via Gromov--WassersteinMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
The Platonic Defense: Backdoor Defense for Self-Supervised Encoders in the Era of Large Scale Pre-trainingLLM safety and security → Backdoor and data poisoning attacks
VeRe-Flow: Guiding Flow Matching toward Clean Speech via Velocity Contrastive Regularization and Representation Alignment for Noise-Robust Bandwidth ExpansionDiffusion models → Audio, speech and music
Bridging VideoQA and Video-Guided Agentic Tasks via Generalized Keyframe ExtractionMultimodal LLMs / vision-language models → Video understanding with MLLMs +1
FADE: Mitigating Hallucinations by Reducing Language-Prior Dominance in Large Vision-Language ModelsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Mixture of Debaters: Learn to Debate at Architectural Level in Multi-Agent ReasoningLLM agents and tool use → Social simulation and multi-agent debate +1
LLM-Guided Planning for Multi-hop Reasoning over Multimodal Nuclear Regulatory DocumentsLLM reasoning → Reasoning for agents, planning and code
NaLA: A 3D Native LLM Layout Agent for High-quality 3D Scene GenerationLLM agents and tool use → Multimodal and embodied agents
Interventional Flow Matching: Prospective Dose-Response Forecasting with Velocity-Field Jacobian RegularizationDiffusion models → Flow matching, time series and science
Event-VLA: Action-Conditioned Event Fusion for Robust Vision-Language-Action ModelVision-language-action models → 3D and spatial-aware VLA
Diagnosing and Repairing Factual Errors in RAG under Budget ConstraintsRetrieval-augmented generation → RAG robustness and adaptive retrieval
L2D2-GS: Learning to Densify for Feedforward Dynamic Gaussian Scene ReconstructionWorld models → World models for autonomous driving
SAFE-DiT: Semantics-Aware Fast-path Execution for High-Resolution Diffusion TransformersDiffusion models → Diffusion foundations and sampling
When LLMs Develop Languages: Symbolic Communication for Efficient Multi-Agent ReasoningLLM reasoning → Reasoning for agents, planning and code +1
Fast Enough to Act: Spatio-Temporal Visual Token Merging for Low-Latency Robotic VLMs and VLAsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +2
PHF: Privileged Hidden Flow for On-Policy Self-DistillationLLM reasoning → Prompting and chain-of-thought reasoning +1
Covering the Unseen: Information Demand Coverage Optimization for Retrieval-Augmented GenerationRetrieval-augmented generation → RAG pipeline design and evaluation
FDM-MFVT: Few-step Sampling Diffusion Model for Mask-Free Virtual Try-OnDiffusion models → 3D, point cloud and scene generation
Hierarchical Experimentalist AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Process Advantage Signal Shaping: A Paradigm-Agnostic Middleware for Process-Supervised RL in LLM ReasonersLLM reasoning → Reinforcement learning for reasoning +2
ASTAD: Asymmetric Style Transfer for Synthetic-to-Real Adaptation in Autonomous DrivingEnd-to-end autonomous driving → Generative simulation for driving
Manufactured Confidence: How Memory Consolidation Turns Hearsay into Confident FactsLLM agents and tool use → Agent memory systems
The Complexity Ceiling Benchmark: A Multi-Domain Evaluation of Sequential Reasoning Under Depth ScalingLLM reasoning → Test-time compute and efficient thinking
Adaptive Block Diffusion: Resolving Training-Inference Mismatch in Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Minority Sentinel: When to Overturn Majority Voting in Multi-Agent LLM DebatesLLM reasoning → Reasoning for agents, planning and code +1
MIThinker: A Plug-and-Play Policy-Optimized Thinker For Motivational Interviewing CounselingLLM reasoning → Clinical and biomedical applications
Travel-Oriented Reasoning Large Language Model via Domain-Specific Knowledge GraphsLLM reasoning → Knowledge- and tool-augmented reasoning
When Summaries Distort Decisions: Information Fidelity in LLM-Compressed Financial AnalysisEfficient LLM inference and serving → Efficient reasoning and agent inference
SurgVLA-Bench: Towards Evaluating Vision-Language-Action Models for Laparoscopic Surgical RoboticsVision-language-action models → Generalist VLA models and benchmarks +1
QuantGuard: Learnable Rounding for Repairing Quantization-Conditioned Backdoors in LLMsLLM safety and security → Backdoor and data poisoning attacks
On the Policy Gradient Foundations of Group Relative Policy Optimization: Credit Assignment, Gradient Sparsity, and Rank CollapseLLM reasoning → Reinforcement learning for reasoning
When Does Synthetic CT Transfer? A Label-Free Donor/Host Diagnostic for Medical Vision-Language Model Routing on Real Lung CTMultimodal LLMs / vision-language models → Medical vision-language models
Understanding Evaluation Illusion in Diffusion Large Language ModelsEfficient LLM inference and serving → Diffusion LLM and MoE inference
PolicyGuard: A Dialogue-Grounded Sub-Agent Verifier for Policy Adherence in LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance
Multi-Block Diffusion Language ModelsDiffusion models → Discrete and text diffusion
Can OCR-VLMs Read Devanagari? A Stress-Test Benchmark and Post-Correction StudyMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
KernelFlume: Elastic Core-Attention Scaling for Agentic Long-Context DecodingEfficient LLM inference and serving → KV cache optimization
DTI: Dynamic Trajectory Initialization for Generative Face Video Super-ResolutionDiffusion models → Inverse problems and image restoration
AI Trading's Alpha Singularity: Emergent Market Reasoning through Agent-to-Agent Self-EvolutionLLM reasoning → Reasoning for agents, planning and code +1
A Multi-Dataset Benchmark for Evaluating LLM Agents in Microservice Failure DiagnosisLLM agents and tool use → Software engineering coding agents
Selective Memory Retention for Long-Horizon LLM AgentsLLM agents and tool use → Agent memory systems
Symbolic Mechanistic Data Attribution: Tracing Training Influence to Learned Behavioral PoliciesMechanistic interpretability → Circuits and mechanistic analysis
Invariant Reasoning Directions in Latent Trajectories of Language ModelsLLM reasoning → Test-time compute and efficient thinking
OASIF: An Efficient Obfuscation-Aware Self-Improving Framework for LLM-Based Assembly Code Instruction Following and ComprehensionSynthetic data and self-improvement for LLMs → Code and instruction data synthesis
Flow Reasoning Models: Turning Flows Into Efficient Recurrent ReasonersLLM reasoning → Diffusion reasoning and test-time scaling +1
Agent Security Meets Regulatory Reality -- A Practitioner Systematization of Autonomous-Agent Threats and Controls in Regulated Financial SystemsLLM agents and tool use → Financial trading and forecasting agents
DistilledGemma: Balanced Efficiency-Accuracy for Person-Place Relation Extraction from Multilingual Historical ArticlesKnowledge distillation → Knowledge distillation for LLMs
Vagdhenu: A Vrutta (Meter) Aware Shloka-to-Chant (TTS) System for Sanskrit
LLMs for Academic Workflows: An Evaluation of Literature Reviews Generated with Short and Long Context Windows of LLMs
Why Current XAI Is Not Enough for Arabic NLP: A Critical Survey of the Explainability Gap
Position: Medical AI Neglects Real Treatment Outcomes
AI-Assisted Knowledge Access for Legacy Enterprise Asset Management in Energy Operations: A Practical Retrieval System
A GAN-Based Framework for Robust Data Synthesis in Satellite Internet Observations
Reducing Per-Sample Interference in Stochastic Optimization
Quantifying Ranking Uncertainty in LLM Benchmarks
Neural Controlled Differential Equations for EMT-Level Surrogate Modeling of Grid-Forming Inverters
Discovery by Dreaming: Cross-Domain Recombination in Artificial Memory
Entropy-Coded MS-VQ-VAE with Learned Priors for Ultra-Low Bitrate Video Compression
Double-Helix Active Geometry: LiDAR-Anchored Multi-View Depth with Selective Abstention

All 253