Papers

New on arXiv

Thu 13 Aug 2026

415 papers

Coding Agents are Strong Prompt OptimizersLLM agents and tool use → Software engineering coding agents
PACE-dLLM: Elastic Block Decoding via Confidence Cliff Estimation for Diffusion Language ModelsDiffusion models → Discrete and text diffusion +1
Machine Unlearning for Speech Question Answering in Large Audio-Language ModelsLLM safety and security → Machine unlearning
Multi-dimensional Bias in Modeling Multi-dimensional Preferences: Evaluating the Ability of Synthetic Agents to Replace Human Participants in Conjoint ExperimentsLLM agents and tool use → Social simulation and multi-agent debate +1
Measuring Collective Semantic Change in Populations of Language Model AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Composable Trust Infrastructure for Manufacturing Knowledge Graphs: Cross-System Provenance, Temporal Reasoning, and Decision TraceabilityLLM reasoning → Reasoning for agents, planning and code
Retrieval-grounded robot program generation and simulation-based correction via Model Context ProtocolLLM agents and tool use → Multimodal and embodied agents
PolyComp: A Polycube-based Benchmark for Compositional 3D Spatial Reasoning in Multimodal ModelsLLM reasoning → Spatial and embodied reasoning
When Lexical Change Misleads: Rethinking Dynamic Topic Model Evaluation with Traditional and LLM-Based MetricsLLM reasoning → Reasoning benchmarks and evaluation
VoiceChat-TTS: A Low-Latency Continuous Speech Synthesis Model for Interactive AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
From Passive Delegates to Strategic Negotiators: Reinforcing Social Reasoning in Small Language Models with SocialRLLLM reasoning → Reasoning for agents, planning and code
ChartProbe: A Diagnostic Study on Visual Reasoning through Perception, Grounding, and Simple ReasoningLLM reasoning → Multimodal reasoning +1
Amplified Does Not Mean Predictive: Reasoning Behaviors in Thinking ModelsLLM reasoning → Test-time compute and efficient thinking
Limitations of Synthetic Data Generation in Specialized Data-Scarce DomainsDiffusion models → Image generation, text-to-image and editing
Graph-MambaNav: Spatial-Temporal Graph Mamba Leveraging Object-Relation Knowledge for Object-Goal NavigationState-space sequence models (S4, Mamba) → Video, motion and generative Mamba models
TeachMateGPT: A Multi-Agent Knowledge-Grounded Framework for Pedagogical Assessment Generation from Science Curriculum MaterialsRetrieval-augmented generation → Agentic RAG, agents and memory
CLAIR-Fin: An Adversarial Multi-Agent Framework for Claim-Level Verification and Adaptive Debate in Cross-Modal Financial QARetrieval-augmented generation → Legal and financial RAG
GRPO Beyond English: A Large-Scale Study of GRPO in Non-English and Multilingual SettingsLLM reasoning → Reinforcement learning for reasoning
hint2: Hierarchical World Models for Inference-Time Temporal Logic GuidanceWorld models → Embodied and robot world models
Multiphase-Diff: Diffusion-Based Generative Modeling for High-Contrast Multiphase Physical Systems with Sharp InterfacesDiffusion models → Flow matching, time series and science
Second Thought: Reasoning in Parallel as LLM Agents Act and ObserveLLM reasoning → Reasoning for agents, planning and code +1
Ontology-Grounded Project Memory for Coding AgentsLLM agents and tool use → Software engineering coding agents
AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic DesignLLM agents and tool use → Multimodal and embodied agents
PlayWorld: Benchmarking World Models with Agent Players over Long-Horizon ObjectivesWorld models → Video-generation world models and simulators
Alaya-EVOKE: From Linear-Scaling Supervision to Endless WorldWorld models → Video-generation world models and simulators
DARTree: Speculative Diffusion Decoding with Autoregressive Draft TreesEfficient LLM inference and serving → Speculative decoding
Intern-S2-Preview: Scientific Agentic Foundation ModelLLM agents and tool use → Web, search, and research agents
OpScale: Operator-level Provisioning and Autoscaling for LLM ServingEfficient LLM inference and serving → Serving systems and request scheduling
AlayaWorld: Interactive Long-Horizon World Modeling - Full Technical Report (v1.1)World models → Video-generation world models and simulators
DreamX-Phi 1.0: Action-Conditioned Video World Model for Robotic ManipulationWorld models → Embodied and robot world models
MapRoute++: Surrogate-Guided Semantic Routing for Visual Concept UnlearningLLM safety and security → Vision-language and text-to-image safety
MARC v1: An Open-Source Multi-Agent Framework for Clinical AI Reasoning and CoordinationLLM reasoning → Clinical and biomedical applications +1
Decoding Task Progress from VLA RepresentationsVision-language-action models → Generalist VLA models and benchmarks
MLLM-Routed Heterogeneous Ensembles for Robust Cross-Dataset Image ClassificationMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
SNM-VFI: Symmetric Nonlinear Motion-Guided Generative Video Frame InterpolationDiffusion models → Video generation
A Unifying Perspective on Causal World Models: From Observations to Representations to StructureWorld models → World models for language agents
UniTexture: Cross-Task Universal Adversarial Textures for Vision-Language-Action ModelsVision-language-action models → Safety and adversarial attacks
Edit2TikZ: A Comprehensive and Challenging Benchmark for Scientific Figure Editing with TikZMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
ContactGuard: Pre-Contact Execution Monitoring with Action-Conditioned Latent World ModelsWorld models → Embodied and robot world models
Enhancing Virtual Agents through SLMs and Edge-Computing: An Exploratory Evaluation of Think and Memory ProcessesLLM reasoning → Spatial and embodied reasoning +1
Who Speaks Matters: Authority-Aware Multi-View RAG over Italian Parliamentary ProceedingsRetrieval-augmented generation → Legal and financial RAG
FIRE-VLA: Failure-Informed Self-Evolution for Vision-Language-Action Models in Autonomous DrivingVision-language-action models → VLA for autonomous driving +1
Rules or Character? Scaling Laws for AI Safety DesignLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
LongEarth-R1: Benchmarking and Aligning Vision-Language Models for Long-Horizon Earth Observation ReasoningLLM reasoning → Spatial and embodied reasoning +1
RippleMem: From Isolated Retrieval to Associative Recollection for Long-Term Agent MemoryLLM agents and tool use → Agent memory systems
Beyond Local Accuracy: A Protocol-Level Identifiability Audit for Controlled LLM Reasoning EvaluationLLM reasoning → Test-time compute and efficient thinking
StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent SystemsLLM agents and tool use → General LLM agent frameworks and benchmarks
Keep, Customize, or Exit: Default Design and Token Pricing in LLM Reasoning ServicesLLM reasoning → Test-time compute and efficient thinking
How Good are Foundation Models in Longitudinal MRI Disease Progression Reasoning?LLM reasoning → Clinical and biomedical applications
Refusing Intent, Not Form: Wrapper-Based Intent-Group Supervision for LLM SafetyLLM safety and security → Safety alignment, refusal and guardrails
Refine After Generation: Toward Correct and Concise Patches in LLM-based Program RepairLLM agents and tool use → Software engineering coding agents
How Do VLMs Behave When Blind or Misled? Behavioral Evaluation of VLMs on Scientific FiguresLLM reasoning → Multimodal reasoning +1
vToken: Token-Level Virtualization for Reclaimable KV CachesEfficient LLM inference and serving → KV cache optimization
Localize, Then Reason: Visual Latent Structural Reasoning for Molecular Properties and EditsLLM reasoning → Clinical and biomedical applications
Reasoning for Social Audio-Visual Question Answering: Where Do We Stand?LLM reasoning → Audio-language model reasoning +1
When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1Retrieval-augmented generation → RAG robustness and adaptive retrieval
CoverPrune: Coverage-Driven Token Pruning for 3D VLMs via Optimal TransportLLM reasoning → Spatial and embodied reasoning +2
TsuGO: Probing Search Efficiency in LLM Reasoning via Go Life-and-Death ProblemsLLM reasoning → Test-time compute and efficient thinking
HPSD: Hybrid-Policy Self-Distillation for Text-Image-to-Video Diffusion ModelsDiffusion models → Video generation
GEM: A Generative Embedding Model Bridging Reasoning and RetrievalLLM reasoning → Knowledge- and tool-augmented reasoning
Teach the Magnitude, Not the Direction: Verifier-Bounded Credit Assignment for Multi-Turn Multi-step LLM AgentsLLM agents and tool use → Reinforcement learning for LLM agents
SkillShapley: Boundary-Adaptive Shapley Valuation for Skill Step Attribution in LLM AgentsLLM agents and tool use → Agent skill learning and evolution
TRAPSBench: Vision-Language Models Encode but Fail to Express Epistemic RestraintMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Better Decomposition, Free Aggregation: A Synthesizer-Folding Framework for Multilingual Multi-Hop Question AnsweringRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
MergeOver: Post-Training Token Merging for Recursive Vision TransformersEfficient LLM inference and serving → Token pruning and merging for ViTs
Numeracy in Large Language Models: Fundamental Limitations and Paths to ImprovementLLM reasoning → Mathematical reasoning
Potential Applications of HBF in LLM Serving SystemsEfficient LLM inference and serving → Serving systems and request scheduling
QuISE: Defense against Typographic Attacks on VLMs via Query-Irrelevant Semantic EditingMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
EgoMonth: A Month-Level Egocentric Video Benchmark for Long-Term Spatiotemporal MemoryMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
S2-HWM: Sparse Event-Structured Hierarchical World Model for Long-Horizon Surgical Robot ManipulationWorld models → Embodied and robot world models
FlowLOB: Efficient and Controllable Limit Order Book Generation with Flow MatchingDiffusion models → Flow matching, time series and science
Semantic Radiance Fields as Simulators for Spatial Reasoning in Real-World ScenesLLM reasoning → Spatial and embodied reasoning
SPADE: Speculative Decoding for Precise and Low Cost Distributed Edge Cloud InferenceEfficient LLM inference and serving → Early exit and edge inference
Behavioral Reprogramming of Open-Weights Models: Cognitive Plasticity and Alignment BoundsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
VALG: An Agentic System for ML Theory ResearchLLM agents and tool use → Educational agents and math problem solving
TEMPO: Makespan-Aware Expert-Parallel Load Balancing Across Memory- and Compute-Bound RegimesMixture-of-experts → Efficient MoE training and inference systems
H2R-Bench: Benchmarking Human-to-Robot Manipulation Video Generation in World ModelsWorld models → Embodied and robot world models
From Local Mismatch to Global Impact: Optimizing Cache Reuse Policy for Efficient DiffusionDiffusion models → Video generation
Spatially-Grounded Text-to-Video Generation via Inference-Time Gradient-Free OptimizationDiffusion models → Video generation
UniTraffic-Agent: Unified Traffic Video Reasoning for AI City Challenge 2026 Track 3 with Two Out-of-Domain EvaluationsLLM reasoning → Chain-of-thought reasoning for driving +1
InterSAGE: The Secure and Verifiable Interoperability Protocol for An Internet of AgentsLLM agents and tool use → Agent threats, guardrails and governance
Static analysis-guided agentic AI translation enables Rust as a full stack bioinformatics languageLLM agents and tool use → Medical and scientific domain agents
Temporal GRPO: Beyond Trajectory-Level Credit in Vision-Language-Action Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning +1
RAGSieve: Self-Referenced Local Contrast for Knowledge-Poison Detection in Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG
HybridRAG-BN: A Retrieval-Augmented Framework with Fine-Tuned Verification for Bangla KBQARetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
PixSDS: Why Latent SDS Makes Noisy PixelsDiffusion models → 3D, point cloud and scene generation
LycheeMemory V2: Efficient Long-Term Memory for LLM Agents via Semantic Segment-Level ConsolidationLLM agents and tool use → Agent memory systems
Reconcile Once, Write Anytime: A Trust-Tiered Librarian and a Multi-Agent Writer for Drift-Free, Point-in-Time ResearchLLM agents and tool use → Web, search, and research agents
Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance +1
The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot UseWorld models → Latent predictive world models
I-SDPO: Instance-Level Adaptive Self-Distillation Policy OptimizationLLM reasoning → Reinforcement learning for reasoning
VoxAudio: Vocalized Audio Synthesis via Multi-Reward Autoregressive Flow MatchingDiffusion models → Audio, speech and music
Diagnosing JEPA World Models with Action-Conditioned Predictive ConsistencyWorld models → Latent predictive world models
FlashDrive: Flash Vision-Language-Action Inference for Autonomous DrivingVision-language-action models → VLA for autonomous driving +2
Polish Medical Visual Question Answering: Vision-Language Models Underutilize Visual EvidenceMultimodal LLMs / vision-language models → Medical vision-language models
Discovering Efficient and Explainable Communication Topologies for LLM-based Multi-Agent Systems via Causal InferenceLLM agents and tool use → General LLM agent frameworks and benchmarks
TennisVAR: A Stroke-Evidence-Grounded Multimodal Large Language Model for Tactical Reasoning in Tennis VideosLLM reasoning → Multimodal reasoning +1
Towards Socially Compliant Navigation in Deep Reinforcement Learning via Proxemics-Based Reward ModelingLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Beyond Visual Evidence: Revealing and Mitigating Relational Privacy Leakage in Document MLLMsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +1
HounsWorld: A Multimodal World Model for Hidden Patient-State Readout, Reconstruction, and SimulationWorld models → Medical and surgical world models

All 415