Papers

New on arXiv

Tue 30 Jun 2026

534 papers

MGDiff: Multi-Interest Sequence Recommendation with Masking GNN-Guided DiffusionDiffusion models → Diffusion recommenders
Multi-Agent Retrieval-Augmented Generation for Efficient Cloud Knowledge Base Search in Telecom SNOC EnvironmentRetrieval-augmented generation → Agentic RAG, agents and memory +1
Hybrid Retrieval-Augmented Generation with Knowledge Graph Expansion, RRF Fusion, and Per-Chunk Grounded Evaluation for Enterprise Document SearchRetrieval-augmented generation → RAG pipeline design and evaluation
Neurosymbolics for Data Engineering: Achieving Long Context Token Reduction Without FinetuningEfficient LLM inference and serving → KV cache optimization
Representation Affects Retrieval: A Case Study of Skill Discovery and Routing in a Multimodal Agent HarnessLLM agents and tool use → Agent skill learning and evolution
EMMR: Emotion-Mediated Multimodal Reasoning for Personality Assessment in Asynchronous Video InterviewsLLM reasoning → Audio-language model reasoning
Reasoning with Memory: A Temporal Granularity-Adaptive Framework for Training-Free Long Video UnderstandingLLM reasoning → Multimodal reasoning +1
Obliviate: Efficient Unlearning in Recommender SystemsLLM safety and security → Membership inference attacks
One Student, Many Teachers: Multi-Task On-Policy Distillation via Soft-Prompt Privileged ContextKnowledge distillation → On-policy and self-distillation for LLMs
Multi-Timescale Latent-Action DRL for Joint Optimization in Edge-Cloud NetworksDeep reinforcement learning methods → RL for energy, networks and operations
Preference-Conditioned Multi-Objective Reinforcement Learning for Runtime-Tunable Transit Signal PriorityDeep reinforcement learning methods → RL for autonomous driving and traffic
Benchmarking the Benchmarks: A Validity Audit of Tool-Calling EvaluationLLM agents and tool use → Tool learning and function calling
Criterion-Conditional In-Context Learning: Evaluating Criterion-Shift Adaptation in Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
From Tensor Buffer to Distributed Memory Hierarchy: A Survey of KV Cache Management for LLM ServingEfficient LLM inference and serving → KV cache optimization
Dual-Adaptive SAM3: Hierarchical Routing over Low-Rank Expert Layers for Parameter-Efficient Medical Image SegmentationMixture-of-experts → Medical and biomedical MoE
Benchmarking Federated Learning and Knowledge Distillation for Point Cloud ClassificationKnowledge distillation → Detection and dense prediction distillation
ASPIRE: Agentic /Skills Discovery for RoboticsLLM agents and tool use → Multimodal and embodied agents
Mnemosyne: Agentic Transaction Processing for Validating and Repairing AI-generated WorkflowsLLM agents and tool use → Software engineering coding agents
From Signals to Structure: How Memory Architecture Drives Language Emergence in LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety GuardsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
ELMP: Efficient Learning for Motion Planning via Analytical Policy GradientsDeep reinforcement learning methods → RL with diffusion and flow models
Play Like Champions: Counterfactual Feedback Generation in Latent SpaceDeep reinforcement learning methods → RL for games and combinatorial optimization
DriftScope: Measuring The Hidden Effects of Diffusion Model AdaptationDiffusion models → Image generation, text-to-image and editing
Steal the Patch Size: Adversarially Manipulate Vision-Language ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
FRAME: Learning the Adaptation Domain with a Mixture of Fractional-Fourier ExpertsMixture-of-experts → Mixture of LoRA experts
Distributed Multi Robot Lunar Cargo Transportation via Phase Decomposed Reinforcement LearningDeep reinforcement learning methods → RL for robot navigation and UAV control
Identifying and Resolving Pitfalls of Knowledge-Based VQA Benchmarks: Auditing, Repairing, and AugmentingMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
SmoothAgent: Efficient Long-Horizon LLM-Based Agent Serving with Lookahead Context EngineeringEfficient LLM inference and serving → Efficient reasoning and agent inference
3D Point World Models: Point Completion Enables More Accurate Dynamics LearningWorld models → Embodied and robot world models
RareDxR1: Autonomous Medical Reasoning for Rare Disease Diagnosis Beyond Human AnnotationLLM reasoning → Clinical and biomedical applications
Decompose, Compare, and Decide: Multimodal LLMs are Implicit Few-Shot LearnersMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
PixelEyes: Decoupling Perception and Reasoning for Pinpoint Visual Evidence SeekingLLM reasoning → Multimodal reasoning
Representation as a Bottleneck for Mechanistic Interpretability: The Manifestation Unit ProtocolMechanistic interpretability → Circuits and mechanistic analysis
Harnessing the Latent Space: From Steering Vectors to Model Calibrators for Control and TrustMechanistic interpretability → Activation steering methods
Synergistic Perception-Reasoning Governance: Grounding Medical MLLMs with Verifiable Anatomical EvidenceLLM reasoning → Hallucination detection and mitigation +1
Joint Medical Image Enhancement and Segmentation with Diffusion-based Symbiotic Information InteractionDiffusion models → Medical imaging and segmentation
SWE-Router: Routing in Multi-turn Agentic Software Engineering TasksLLM agents and tool use → Software engineering coding agents
AGE: Adaptive-masking for Graph Embedding in Graph Retrieval-Augmented GenerationRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
FaceMoE: Mixture of Experts for Low-Resolution Face RecognitionMixture-of-experts → Vision mixture-of-experts
QVal: Cheaply Evaluating Dense Supervision Signals for Long-Horizon LLM AgentsLLM agents and tool use → Reinforcement learning for LLM agents
SpheRoPE: Zero-Shot Optimization-Free 360 Panorama Generation with Spherical RoPEDiffusion models → 3D, point cloud and scene generation
Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
DVG-WM: Disentangled Video Generation Enables Efficient Embodied World Model for Robotic ManipulationWorld models → Embodied and robot world models +1
AdaJEPA: An Adaptive Latent World ModelWorld models → Latent predictive world models
Generative Skill Composition for LLM AgentsLLM agents and tool use → Agent skill learning and evolution
Cross-Space Distillation: Teaching One-Step Students with Modern Diffusion TeachersDiffusion models → Diffusion foundations and sampling +1
TRIAGE: Role-Typed Credit Assignment for Agentic Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents
Surrogate Fidelity: When Can Open LLMs Explain Closed Ones?Mechanistic interpretability → Knowledge and representations inside LLMs
AxDafny: Agentic Verified Code Generation in DafnyLLM agents and tool use → Software engineering coding agents
On the Comparison of Reinforcement Learning and Adaptive Control for Linear Systems under Packet Loss and UncertaintyDeep reinforcement learning methods → RL for games and combinatorial optimization
Self-Study Reconsidered: The Hidden Fragility of Learning from Self-Generated QASynthetic data and self-improvement for LLMs → Synthetic reasoning data and bootstrapping
CoLT: Teaching Multi-Modal Models to Think with Chain of Latent ThoughtsLLM reasoning → Multimodal reasoning +1
GR2 Technical ReportKnowledge distillation → Distillation for recommender systems
ERA: Entropy-Guided Visual Token Pruning with Rectified Attention for Efficient MLLMsMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
DigitalCoach: Communication and Grounding Gaps in Human and Agentic Computer Use CoachingLLM agents and tool use → General LLM agent frameworks and benchmarks
TreeAgent: A Generalizable Multi-Agent Framework for Automated Bias Labeling in Forestry via Compiled Expert Rules and Vision-Language ModelsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
MECoBench: A Systematic Study of Multimodal Agent Collaboration in Embodied EnvironmentsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
Signed-Permutation Coordinate Transport for RMSNorm TransformersMechanistic interpretability → Circuits and mechanistic analysis
Adapting Generalist Robot Policies with Semantic Reinforcement LearningVision-language-action models → Reinforcement learning and post-training +1
DriveWeaver: Point-Conditioned Video Inpainting for Controllable Vehicle Insertion in Autonomous Driving SimulationWorld models → World models for autonomous driving
Theory of Mind and Persuasion Beyond Conversation: Assessing the Capacity of LLMs to Induce Belief States via Planning and ActionLLM agents and tool use → General LLM agent frameworks and benchmarks
Attend, Transform, or Silence: Operator-Level Visual Skipping for Efficient Multimodal LLM InferenceMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Z-1: Efficient Reinforcement Learning for Vision-Language-Action ModelsVision-language-action models → Reinforcement learning and post-training +1
An Agentic AI Framework to Accelerate Scientific Discovery in Plant PhenotypingLLM agents and tool use → Web, search, and research agents
PriorEye: Geospatial Visual Priors for End-to-End Autonomous DrivingEnd-to-end autonomous driving → End-to-end system architecture
Breaking Failure Cascades: Step-Aware Reinforcement Learning for Medical Multimodal ReasoningLLM reasoning → Clinical and biomedical applications +1
Reinforcement Learning-Based Control for an Inline Skating Humanoid RobotDeep reinforcement learning methods → RL for robot locomotion and manipulation
Evo-PI: Aligning Medical Reasoning via Evolving Principle-Guided SupervisionLLM reasoning → Clinical and biomedical applications +2
Bridging the Gap Between Latent and Explicit Reasoning with Looped TransformersLLM reasoning → Test-time compute and efficient thinking
Mesh BDF: Barycentric Dominance Field for 3D Native Mesh GenerationDiffusion models → 3D, point cloud and scene generation
A Self-Evolving Agentic System for Automated Generation and Execution of Biological ProtocolsLLM agents and tool use → Medical and scientific domain agents
Addressing Over-Refusal in LLMs with Competing RewardsLLM safety and security → Safety alignment, refusal and guardrails
Rhythm-Structured Predictive Learning for Remote PhotoplethysmographyState-space sequence models (S4, Mamba) → Physiological signal modeling
MemLearner: Learning to Query Context memory for Video World ModelsWorld models → Video-generation world models and simulators
UniCoder: Unified Visual-to-Code Generation via Symbolic Rewards and Reference-Guided Code OptimizationMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
UniTacVLA: Unified Tactile Understanding and Prediction in Vision Language Action ModelsVision-language-action models → Tactile, force-aware and humanoid VLA
Seeing Is Not Sharing: Some Vision-Language Models Overestimate Common Ground in Asymmetric DialogueMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Diffusing Blame: Task-Dependent Credit Assignment in Biologically Plausible Dual-Stream NetworksDeep reinforcement learning methods → Deep RL algorithms and representations
Look But Don't Touch with Sparse Autoencoders for Unlearning in Diffusion ModelsDiffusion models → Diffusion foundations and sampling +2
ShopX: A Foundation Model for Intent-to-Item Fulfillment in Agentic ShoppingLLM agents and tool use → Social simulation and multi-agent debate
FastDSAC: Enhancing Policy Plasticity via Constrained Exploration for Scalable Humanoid LocomotionDeep reinforcement learning methods → RL for robot locomotion and manipulation
Histogram-constrained Image GenerationDiffusion models → Image generation, text-to-image and editing
WorldRoamBench: An Open-World Benchmark for Long-Horizon Stability of Interactive World ModelsWorld models → Video-generation world models and simulators
SAMBA: A Scatter-Guided Masked Bidirectional Mamba Foundation Model for SAR Target RecognitionState-space sequence models (S4, Mamba) → Point clouds, detection and remote sensing
ForecastAgentSearch: Towards a Multi-Expert Agent Search System for Geopolitical Event ForecastingLLM agents and tool use → Financial trading and forecasting agents
FARS: A Fully Automated Research System Deployed at ScaleLLM agents and tool use → Web, search, and research agents
ECHO: Prune To Act, Trace To Learn With Selective Turn Memory In Agentic RLLLM agents and tool use → Agent memory systems
Think in English, Answer in Korean: Efficient Adaptation of Multilingual Tool-Using AgentsLLM reasoning → Reasoning for agents, planning and code +1
Technical Report of RoboSpatial Challenge at CVPR 2026: Selective Reasoning Activation and Reference-Frame Disambiguation for Embodied Spatial ReasoningLLM reasoning → Spatial and embodied reasoning +1
A Tutorial on Autonomous Fault-Tolerant Control Using Knowledge-Grounded LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
What Memory Do GUI Agents Really Need? From Passive Records to Active Task-Driving StatesLLM agents and tool use → GUI and mobile agents
CLExEval: A Human-in-the-Loop Framework for Qualitative Evaluation of LLM Clinical ReasoningLLM reasoning → Clinical and biomedical applications
Preserve the Hard, Regenerate the Rest: Uncertainty-Guided Synthetic Training Data Augmentation with Diffusion ModelsDiffusion models → Medical imaging and segmentation
Robust Text Watermarking for Large Language Models via Dual Semantic EmbeddingsLLM safety and security → Watermarking generated content and models
Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement LearningLLM reasoning → Clinical and biomedical applications +2
LASER: Load-Aware Serving with Early-Exit for Reasoning LLMs at the EdgeEfficient LLM inference and serving → Serving systems and request scheduling
DualBrep: A Dual-Field Continuous Representation for B-rep ModellingDiffusion models → 3D, point cloud and scene generation
Localized Conformal Prediction for Image Classification with Vision-Language ModelsMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Introduction to Stochastic Differential Equations for Generative Machine Learning: A Variational PerspectiveDiffusion models → Diffusion foundations and sampling
Which Tokens Matter? Adaptive Token Selection for RLVR with the Relative Surprisal IndexLLM reasoning → Reinforcement learning for reasoning

All 534