Papers

New on arXiv

Sun 21 Jun 2026

196 papers

The Race between Agentic AI Capabilities and Data Quality Control in Online SurveysLLM agents and tool use → Web, search, and research agents
Wazobia Eval: A Benchmark for Nigerian Pidgin Emotion Understanding, Sarcasm Detection, and Cultural ReasoningLLM reasoning → Reasoning benchmarks and evaluation
Trajectory Learning with Graph Representations for Social Robot NavigationOffline RL and imitation learning from data → Social robot navigation
Emergent Culture in Minimal LLM SystemsLLM agents and tool use → General LLM agent frameworks and benchmarks
Zero-Label Driving Scenario Complexity Detection via Joint Embedding Predictive ArchitectureEnd-to-end autonomous driving → VLMs for autonomous driving
Sol Video Inference Engine: Agent-Native Full-Stack Acceleration Framework for Efficient Video GenerationDiffusion models → Video generation
A Survey on Federated Causal Discovery and InferenceLLM reasoning → Causal reasoning in LLMs
Weight-Space Geometry of Offline Reasoning TrainingLLM reasoning → Reinforcement learning for reasoning +1
Systematic Exploration of 4-Expert Heterogeneous Mixture-of-Experts via Automated Pipeline SearchMixture-of-experts → MoE theory and gating functions
Habituation at the Gate: Rising Approval and Declining Scrutiny in Human Review of AI Agent CodeLLM agents and tool use → Software engineering coding agents
One-Prompt Censorship Evasion via Generative Diffusion ModelsDiffusion models → Robustness, privacy and adversarial aspects
Beyond Penalizing Mistakes: Stabilizing Efficiency Training in Large Reasoning Models via Adaptive Correct-Only RewardsLLM reasoning → Test-time compute and efficient thinking
Beyond Simpson's Paradox: A Cascade of Confounders in AI Agent Pull-Request Co-AuthorshipLLM agents and tool use → Software engineering coding agents
Libretto: Giving LLM Agents a Sense of Musical StructureLLM agents and tool use → Multimodal and embodied agents
Modular Diffusion Models for Structured Visual RecognitionDiffusion models → Image generation, text-to-image and editing
Black-Box Forensics for Conversational LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance
SATURN: Symbolic Spatial Reasoning for Multi-Perspective GroundingLLM reasoning → Spatial and embodied reasoning +1
The Geometry of Refusal: Linear Instability in Safety-Aligned LLMsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Only Ask What You Don't Know: Grounded Delta Planning for Efficient Multi-step RAGLLM reasoning → Knowledge- and tool-augmented reasoning +1
RigorBench: Benchmarking Engineering Process Discipline in Autonomous AI Coding AgentsLLM agents and tool use → Software engineering coding agents
AgentLens: Interpretable Safety Steering via Mechanistic Subspaces for Multi-Turn Coding AgentLLM agents and tool use → Agent threats, guardrails and governance +1
Prompting Diffusion Models for Zero-Shot Instance SegmentationDiffusion models → Medical imaging and segmentation
A Markov Chain Approach to Preference AlignmentLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
RAVEN: Agentic RAG for Automated Vulnerability RepairLLM agents and tool use → Agent threats, guardrails and governance
Scalable Maximum Entropy Reinforcement Learning for Diffusion Policies via Adjoint MatchingDiffusion models → Diffusion and flow policies for robotics
DR-Mamba: Automatic Inference-Time Domain Adaptation for Document Image Binarization via Sample-Conditioned Detail-Background SuppressionState-space sequence models (S4, Mamba) → Mamba for image restoration
OmniSpace: Efficient Geometry Awareness for Autonomous Vehicles MLLMsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
SkillAudit: From Fixed-Suite Benchmarking to Skill-Centered AssessmentLLM agents and tool use → Agent skill learning and evolution
PaperClaw: Harnessing Agents for Autonomous Research and Human-in-the-Loop RefinementLLM agents and tool use → Web, search, and research agents
MapReason-OSM: Can Vision-Language Models Make Graph-Verifiable Mobility Decisions from Street Maps ?LLM reasoning → Spatial and embodied reasoning +1
Bridging Self-Supervised Learning and Speech Enhancement: A Wav2Vec2-Conditioned FrameworkDiffusion models → Audio, speech and music
What are Key Factors for Updates in RL for LLM Reasoning?LLM reasoning → Reinforcement learning for reasoning
SeFi-Image: A Text-to-Image Foundation Model with Semantic-First DiffusionDiffusion models → Image generation, text-to-image and editing
Look Light, Think Heavy: What Multimodal Chain-of-Thought Reasoning Can and Cannot DoLLM reasoning → Multimodal reasoning
MacAgentBench: Benchmarking AI Agents on Real-World macOS DesktopLLM agents and tool use → GUI and mobile agents
ASAP: A Disaggregated and Asynchronous Inference System for MoE PrefillMixture-of-experts → Efficient MoE training and inference systems
PolicyTrim: Boosting Intrinsic Policy Efficiency of Vision-Language-Action ModelsVision-language-action models → Reinforcement learning and post-training
NegAS: Negative Label Guided Attention and Scoring for Out-of-Distribution Object Detection with Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Governance Decay: How Context Compaction Silently Erases Safety Constraints in Long-Horizon LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance
Robust Diffusion Models via Divergence-Induced Weighted DenoisingDiffusion models → Diffusion foundations and sampling
Imagine to Ensure Safety in Hierarchical Reinforcement LearningWorld models → Latent dynamics models and model-based RL
Lingering Authority: Revocable Resource-and-Effect Capabilities for Coding AgentsLLM agents and tool use → Software engineering coding agents
Benchmarking Vision-Language Models for Microscopic Plant Image UnderstandingMultimodal LLMs / vision-language models → Agricultural vision-language models
Grounded Scaling: Why Agentic AI Needs Deterministic EnvironmentsLLM agents and tool use → General LLM agent frameworks and benchmarks
SCOPE: Evolving Symbolic World for Planning in Open-Ended EnvironmentsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
VADAOrchestra: Neurosymbolic Orchestration of Adaptive Reasoning WorkflowsLLM reasoning → Reasoning for agents, planning and code
Governed AI-Assisted Engineering: Graduated Human Oversight for Agentic Code Generation in Regulated DomainsLLM agents and tool use → Agent threats, guardrails and governance
Lighting-Consistent Object Transfer Across Radiance FieldsDiffusion models → 3D, point cloud and scene generation
CVSBench: A Comprehensive Benchmark for Cross-view Spatial Reasoning and DreamingLLM reasoning → Spatial and embodied reasoning +1
Self-Evolving Cognitive Framework via Causal World Modeling for Embodied Scientific IntelligenceWorld models → World models for language agents
DreamUV: Unwrap Artist-like UV by End-to-End Flow MatchingDiffusion models → 3D, point cloud and scene generation
Words as Difference Makers: How Large Language Models Determine Causal Structure in TextWorld models → World models for language agents
Knowledge-Graph Grounding Helps LLMs Only for Out-of-Training Knowledge: A Controlled Study on Clinical Question AnsweringRetrieval-augmented generation → Medical and biomedical RAG
Code Isn't Memory: A Structural Codebase Index Inside a Coding AgentLLM agents and tool use → Software engineering coding agents
Gold Points Sniper: Self-guided Visual Reasoning in VLM for Fine-grained Action UnderstandingLLM reasoning → Spatial and embodied reasoning
Curvature-Adaptive Consistency Flow Matching: Autonomous Trajectory Optimization via Reinforcement LearningDiffusion models → Diffusion and flow policies for robotics
PlanBench-XL: Evaluating Long-Horizon Planning of LLM Tool-Use Agents in Large-Scale Tool EcosystemsLLM agents and tool use → Tool learning and function calling
MetaPS: Adaptive Programmatic Strategy Selection for Market AgentsLLM agents and tool use → Financial trading and forecasting agents
ARIA: A Causal-Aware Framework for Rescuing LLM Reasoning in Trustworthy Materials DiscoveryLLM reasoning → Clinical and biomedical applications
ZeroGVC: Zero-Shot Generative Video Compression with Autoregressive Diffusion PriorsDiffusion models → Diffusion for compression and wireless
Towards Error-Free Long Video GenerationDiffusion models → Video generation
Reference-Free Assessment of Physical Consistency in World Model-based Video GenerationWorld models → Latent predictive world models
T-IMPACT: A Severity-Aware Benchmark for Contextual Image-Text ManipulationMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Hypothesis-Driven Skill Optimization for LLM AgentsLLM agents and tool use → Agent skill learning and evolution
BabelJudge: Measuring LLM-as-a-Judge Reliability Across Languages and Agent TrajectoriesLLM agents and tool use → General LLM agent frameworks and benchmarks
Geometry-Aware Online Scheduling for LLM Serving: From Theoretical Bound to System PracticeEfficient LLM inference and serving → Serving systems and request scheduling
Curriculum Reinforcement Learning Can Incentivize Reasoning Capacity in LLMs Beyond the Base ModelLLM reasoning → Reinforcement learning for reasoning
Learning at the Right Pace: Adaptive Data Scheduling Improves LLM Reinforcement LearningLLM reasoning → Reinforcement learning for reasoning
FlowDPG: Deterministic Policy Gradient on Flow Matching Policies for Real-World ManipulationDiffusion models → Diffusion and flow policies for robotics
Which India Survives Translation? Narrative Homogenisation Across Indian Oral Traditions in LLMs
The Multilingual Quantization Tax: Structural Collapse and Typological Fragility in Edge SLMs
Do Methods Support the Claims? Intra-Paper Verification for Peer Review
Personalization, Personas, and Forecasting in Value Alignment
Physics-Informed Domain-Invariant Feature Learning with Autoencoder-Driven Gaussian Clustering for Robust Non-line-of-Sight Scenarios
Urban Deceleration Behavior Modes Under Scene Context: An Early-Kinematic Classifier from Argoverse 2 Multi-Agent Trajectories
FLYNN: Robust Neural Network for Robot Navigation using Fly Brain Topology
Neural Speaker Diarization via Multilingual Training: Evaluation on Low-Resource Nepali-Hindi Speech
JEDEL: Zero-Shot DNA-Encoded Library Design for Early-Stage Drug Discovery
Performance and Interpretability of Convolutional, Transformer, and Hybrid Deep Learning Models in Colorectal Histology Classification
Low-power analogue neural networks with trainable nonlinear connections for continuous control
Interpretable Uncertainty Routing Separating Emotion Ambiguity from Distribution Shift in Facial Expression Recognition
Subspace-Constrained Federated Learning with Low-Rank Adaptation
BLUEX v2: Benchmarking LLMs on Open-Ended Questions from Brazilian University Entrance Exams
moBERTo: A Modern Encoder for Portuguese via Continued Pretraining of ModernBERT
Leakage-Aware Benchmarking of LLM Forecasting: Real-Time Nowcasts as the Decision-Time Input for Macro Factor Ranking
Generative Relightable Avatars
Safety-Aware Evaluation of LLM-Generated Driver Intervention Messages through Multi-Task Risk Fusion
SCRUB-FL: Sanitizing and Cleansing Representations via Unlearning of Backdoors
Catching Lies Without Sending the Video: Privacy-Preserving Multimodal Deception Detection
NullFlow: One-Step Generative Reconstruction
VISTA Architect: A graph database-oriented health AI system demonstrated in multidisciplinary tumor boards
GARIP: A Running-Average Moving Reference for Last-Iterate Self-Play in Two-Player Zero-Sum Games
Skin-Deep: A Geometric Diagnostic for Alignment Fragility in Large Language Model Representations
Data Evolution by Wittgenstein's Rule Following
Clipping the Price of Adaptivity at the Tail
From Complaint Narratives to Monetary Relief: A Hybrid Machine Learning Framework for CFPB Consumer Complaints
LSTM Variants for Chaotic Dynamical Systems: An Empirical Study on the Lorenz Attractor
Confidently Wrong: Severity-Aware Calibration of Prompt-Injection Detectors under Attack Shift
Foundation Models for Epileptogenic Zone Identification in Drug-Resistant Epilepsy
MaRS: Robust Out-of-Distribution Detection via Mahalanobis Residual Scoring

All 196