Papers

New on arXiv

Tue 14 Jul 2026

371 papers

CHSR-RRF: A curriculum-gated hybrid retrieval framework with reciprocal rank fusion and leakage-aware benchmarking for educational RAGRetrieval-augmented generation → RAG robustness and adaptive retrieval
Risk-Aware Decision-Making for Autonomous Overtaking: A World Model-Based Mixture-of-Experts FrameworkWorld models → World models for autonomous driving +1
Beyond the Chatbot: Co-Learning and Co-Teaching through a Dual-Persona Generative-AI AssistantRetrieval-augmented generation → Agentic RAG, agents and memory
AutoMem: A Text-Gradient Recursive Self-Improvement Framework for Automated Memory Architectures SearchLLM agents and tool use → Agent memory systems +1
Agreement Is Not Alignment: Divergent Moral Grounds in Human and LLM Ethical JudgmentsLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-ExpertsMixture-of-experts → Expert routing strategies +1
Width, Memory, and Delay: A Resource Accounting for the Limits of Flat Multi-Agent SystemsLLM agents and tool use → General LLM agent frameworks and benchmarks
Motif-Mamba: network motif improved mamba for long-range sequence modelingState-space sequence models (S4, Mamba) → Core SSM architectures for language
FinCacheServe: Dependency-Consistent Answer Reuse for Cost-Efficient RAG Serving over Mutable Enterprise DocumentsRetrieval-augmented generation → Generation and context integration
Seeing What Is Actually There: PriVE-Bench and PriVE-Tools for Counterfactual Evaluation of Agentic Visual Evidence in VLMsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning +1
CoBind: Stage-Aware Compositional Binding for Training-Free Text-to-Image GenerationDiffusion models → Image generation, text-to-image and editing
LookME: Lookup-Based Multimodal Embeddings for Layer Injection in Vision-Language ModelsMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning +1
Med-OPD: Improving Medical Vision-Language Models via Evidence-Aware On-Policy DistillationLLM reasoning → Clinical and biomedical applications +2
MemoHarness: Agent Harnesses That Learn from ExperienceLLM agents and tool use → Software engineering coding agents
Orchestrating Power Grid Studies with Multi-Agent AI and MCP ServersLLM agents and tool use → Financial trading and forecasting agents
Enhancing Small Language Models Reasoning through Knowledge Graph GroundingLLM reasoning → Knowledge- and tool-augmented reasoning +1
Breaking Refusal in the First Half: A Mechanistic Study of the Prefill JailbreakLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
ToolAnchor: Anchoring Counterfactual Context to Boost Agentic Tool-use CapabilityLLM agents and tool use → Tool learning and function calling
Efficient Text-to-Audio Generation via PruningDiffusion models → Audio, speech and music
Meta-Learning Preferences for Multilingual LLM AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Improving Medical Image Generative Models with Fr'echet Distance LossDiffusion models → Medical imaging and segmentation
Harness Handbook: Making Evolving Agent Harnesses Readable,Navigable, and EditableLLM agents and tool use → Software engineering coding agents
Adapting Diffusion-Based Music Synthesis to Speech and Singing Voice ConversionDiffusion models → Audio, speech and music
Deconstructing Actor-Critic: A Large-scale Empirical Study of Design Components for PractitionersDeep reinforcement learning methods → RL theory and sample complexity
Cognitive Commons in the Age of Generative Intelligence: A Heterodox Appraisal of the Knowledge Erosion HypothesisLLM agents and tool use → General LLM agent frameworks and benchmarks
AffectFlow-DINO: Uncertainty-Aware Multi-Task Affect Estimation via Conditional Rectified FlowDiffusion models → Audio, speech and music
GSM-Plus-BN: A Perturbation-Based Benchmark for Bangla Mathematical Reasoning in Large Language ModelsLLM reasoning → Mathematical reasoning
Networked Intelligence: Active Shared Context Graphs for Human-AI Team ScienceLLM reasoning → Knowledge- and tool-augmented reasoning +1
Adaptive Filtering of the KV Cache: Diagnosing and Correcting Structural-Role Bias in LLM InferenceEfficient LLM inference and serving → KV cache optimization
From Human-Centric to Agentic Code Review: The Impact of Different Generations of Generative AI Technology on Review QualityLLM agents and tool use → Software engineering coding agents
RAGthoven at SemEval-2026 Task 1: A Multi-Stage Pipeline Walks Into a Benchmark and Barely Clears the BarLLM agents and tool use → Multimodal and embodied agents
SteinGate: Tail-Sensitive Safe Reinforcement Learning via Stein DiscrepancyDeep reinforcement learning methods → Safe and constrained RL
Learning Safe Agent Behaviour from Human Preferences and Justifications via World ModelsWorld models → Latent dynamics models and model-based RL
Text2Sign: A Single-GPU Diffusion Baseline for Text-to-Sign Language Video GenerationDiffusion models → Sign language production
Boogu-Image-0.1: Boosting Open Agentic Multimodal Generation via Understanding under a Minimal BudgetLLM agents and tool use → Multimodal and embodied agents
ShortOPD: Recovering Pruned LLMs with Short-to-Long On-Policy DistillationKnowledge distillation → On-policy and self-distillation for LLMs
CoDiffGRN: Rethinking Gene Regulatory Network Inference via the BEELINE-KGC Benchmark and Co-evolutionary Discrete DiffusionDiffusion models → Flow matching, time series and science
Improving Molecular Property Prediction in Small Language Models Using Graph-based ToolsLLM agents and tool use → Medical and scientific domain agents
A Hybrid Mamba for Audio-Visual NavigationState-space sequence models (S4, Mamba) → Speech and audio SSMs
Self-Improvements in Modern Agentic Systems: A SurveyLLM agents and tool use → General LLM agent frameworks and benchmarks +1
WaterMoE: Expert-Routing-based Watermarking for High Fidelity and EfficiencyMixture-of-experts → Expert routing strategies
Full-Pipeline Inference Optimization for MiMo-V2.5 Series: Pushing Hybrid SWA Efficiency to the LimitMixture-of-experts → Efficient MoE training and inference systems +1
Efficient and Privacy Aware Edge Cloud Collaborative Inference for Large Language ModelsEfficient LLM inference and serving → Early exit and edge inference
Do AI Agents Know When a Task Is Simple? Toward Complexity-Aware Reasoning and ExecutionLLM reasoning → Reasoning for agents, planning and code +1
DenseReward: Dense Reward Learning via Failure Synthesis for Robotic ManipulationOffline RL and imitation learning from data → Language-conditioned policies and VLAs +1
The Seriality Gap in Video Diffusion ModelsDiffusion models → Video generation
TerraZero: Procedural Driving Simulation for Zero-Demonstration Self-Play at ScaleEnd-to-end autonomous driving → Imitation-learning end-to-end planners +1
PalmClaw: A Native On-Device Agent Framework for Mobile PhonesLLM agents and tool use → GUI and mobile agents
A Shortcut to Statistically Steady-State Turbulence with Flow MatchingDiffusion models → Flow matching, time series and science
FlowWAM: Optical Flow as a Unified Action Representation for World Action ModelsWorld models → Embodied and robot world models
Audio-Native Speech Recognition with a Frozen Discrete-Diffusion Language ModelDiffusion models → Audio, speech and music
Watermark Forensics for Generative Models: An Information-Theoretic PerspectiveLLM safety and security → Watermarking generated content and models
ChunkFlow: Towards Continuity-Consistent Chunked Policy LearningVision-language-action models → Efficient VLA inference and adaptation
Resist and Update: Counterfactual Report Coordinates for Incentive-Compatible LLMsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
FormalAnalyticGeo: A Neural-Symbolic Based Framework for Multimodal Analytic Geometry Problem GenerationLLM reasoning → Spatial and embodied reasoning +1
Exact and Calibrated Diffusion Reconstruction for Digital Breast TomosynthesisDiffusion models → Medical imaging and segmentation
ExToken: Structured Exploration for Efficient Vision-Language-Action Reinforcement Fine-tuningVision-language-action models → Reinforcement learning and post-training +1
Knowledge- and Gradient-Guided Reinforcement Learning for Parametrized Action Markov Decision ProcessesDeep reinforcement learning methods → Interpretable and logic-guided RL
Open-KNEAD: Knowledge-grounded Nutrition Estimation via Agentic DecompositionMultimodal LLMs / vision-language models → Document understanding and multimodal RAG +1
Hy-Embodied-VLM-1.0: Efficient Physical-World AgentsLLM agents and tool use → Multimodal and embodied agents
MemOps: Benchmarking Lifecycle Memory Operations in Long-Horizon ConversationsLLM agents and tool use → Agent memory systems
Unveiling Complex Collective Behaviors from Simple RewardsDeep reinforcement learning methods → RL for robot navigation and UAV control
Verifier-Based Reinforcement Fine-Tuning of Reasoning Models for Thermal Energy Storage ControlLLM reasoning → Reinforcement learning for reasoning
Can LLMs Write Reliable Rubrics? A Meta-Evaluation for Experiment ReproductionLLM agents and tool use → Web, search, and research agents
Accelerating Masked Diffusion Large Language Models: A Survey of Efficient Inference TechniquesDiffusion models → Discrete and text diffusion
Breaking D'ej`a Vu: Independent Auditing of Visual Place Recognition through Vision-Language ReasoningLLM reasoning → Spatial and embodied reasoning
Visual Access Boundaries in Vision-Language Model ReasoningLLM reasoning → Multimodal reasoning +1
UniVR: Thinking in Visual Space for Unified Visual ReasoningLLM reasoning → Multimodal reasoning
Silent Alarm: A J-Space Protocol for Comparing Danger Recognition Across Models and Quantization LevelsLLM safety and security → Safety alignment, refusal and guardrails
Who Grades the Grader? Co-Evolving Evaluation Metrics and Skills for Self-Improving LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks +1
Do We Really Need Multimodal Emotion Language Models Larger Than 1B Parameters?Multimodal LLMs / vision-language models → Multimodal emotion understanding
CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language ModelsLLM reasoning → Multimodal reasoning +1
Directional Constraints for Efficient Exploration in Safe Reinforcement LearningDeep reinforcement learning methods → Safe and constrained RL
Learning Mechanistic Reasoning for Chemical Reactions with Large Language ModelsLLM reasoning → Clinical and biomedical applications
EvoGraph-R1: Self-Evolving Multimodal Knowledge Hypergraphs for Agentic RetrievalRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG +2
Constraint-Aware Aggregation for Federated Reinforcement Learning in Microgrid Energy CoordinationDeep reinforcement learning methods → RL for energy, networks and operations
VisCo: Leveraging Large Language Models as Intrinsic Encoders for Visual Token CompressionMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
RFMSR: Residual Flow Matching for Image Super-ResolutionDiffusion models → Inverse problems and image restoration
Tracing Agentic Failure from the Flow of SuccessLLM agents and tool use → Software engineering coding agents
LLMs Can See the Smoke but not the Fire: Evaluating Abductive Reasoning with ElenchosLLM reasoning → Reasoning benchmarks and evaluation
MaxSAT-Based Feedback for Guiding Vision-Language Models in SudokuLLM reasoning → Reasoning benchmarks and evaluation +1
Less Experts, Faster Decoding: Cost-Aware Speculative Decoding for Mixture-of-ExpertsMixture-of-experts → Efficient MoE training and inference systems +1
MambaPSA: A Mamba-based Replacement for C2PSA in YOLO26State-space sequence models (S4, Mamba) → Point clouds, detection and remote sensing
ReflectVLN: Training Vision-Language Navigation Agents with Reflective ReasoningLLM agents and tool use → Multimodal and embodied agents
MAGE: Color-Invariant and Spatial Knowledge Distillation for Gastric Neoplasm ClassificationKnowledge distillation → Distillation for medical imaging
Jetson-PI: Towards Onboard Real-Time Robot Control via Foresight-Aligned Asynchronous InferenceVision-language-action models → Efficient VLA inference and adaptation
Evidence-Grounded Verified Agentic Reasoning: A Path Toward Eliminating LLM Hallucination in Empirical Inference via Tool-Attested Kernel ProofsLLM reasoning → Reasoning benchmarks and evaluation +1
A Learning-Rate-Gated Failure of GRPO in a Small Language and Vision-Language Model Web Agent: A Controlled Null and Its MechanismLLM reasoning → Reinforcement learning for reasoning +2
Can Induced Emotion Bias LLM Behaviors in Sequential Decision Making?LLM agents and tool use → Social simulation and multi-agent debate
Gradient-free learning of a closed-loop wall controller for turbulent drag reductionDeep reinforcement learning methods → RL with diffusion and flow models
KnowAct-GUIClaw: Know Deeply, Act Perfectly, Personal GUI Assistant with Self-Evolving Memory and SkillSynthetic data and self-improvement for LLMs → Self-evolving LLM agents
PVDetector: Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents through Policy-Violation Concept AnalysisLLM agents and tool use → Agent threats, guardrails and governance +1
Agentic Service-Oriented Computing: A Manifesto for the Next Frontier of Service-Oriented ComputingLLM agents and tool use → General LLM agent frameworks and benchmarks
The Geometry of Memorization: Finite-Time Spectral Sensitivity as a Diagnostic for Flow Matching ModelsDiffusion models → Flow matching, time series and science
Multi-Perspective Agentic Program Repair via Code Property Graphs and Temporal Execution GraphsLLM agents and tool use → Software engineering coding agents
Environment Parameter Gradient Theorem for Co-Design in Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
UD-ASD: A Unified Diffusion Model for Anomalous Sound DetectionDiffusion models → Audio, speech and music
How Agentic Is Agentic Commerce? A Population-Scale Measurement of x402 Adoption and AuthenticityLLM agents and tool use → Agent threats, guardrails and governance
TrustVLA: Mechanism-Guided Inference-Time Defense Against Vision-Language-Action BackdoorsVision-language-action models → Safety and adversarial attacks
Gaussian Mixture Modeling for Event-Aware Visual Allocation in Long Video UnderstandingMultimodal LLMs / vision-language models → Video understanding with MLLMs

All 371