Papers

New on arXiv

Sun 26 Apr 2026

199 papers

SDE Guided Monte Carlo Reinforcement Learning: A Stochastic Maximum Principle Approach for Robust Decision Making in Noisy EnvironmentsDeep reinforcement learning methods → RL theory and sample complexity
Infrastructure for the Agentic Web: Gap Analysis and Architecture from the Agentverse PlatformLLM agents and tool use → General LLM agent frameworks and benchmarks
DeepSeek-V4: Towards Highly Efficient Million-Token Context IntelligenceMixture-of-experts → MoE large language models
Benchmarking Web Agent Safety under E-commerce Deceptive InterfacesLLM agents and tool use → Agent threats, guardrails and governance
The Interlocutor Effect: Why LLMs Leak More Personal Data to Agents Than HumansLLM agents and tool use → Agent threats, guardrails and governance
Distributional Reinforcement Learning via the Cram'er DistanceDeep reinforcement learning methods → RL theory and sample complexity
Adversarial Flow Matching for Imperceptible Attacks on End-to-End Autonomous DrivingDiffusion models → Diffusion and flow policies for robotics +1
Generative Synthetic Data for Causal Inference: Pitfalls, Remedies, and OpportunitiesSynthetic data and self-improvement for LLMs → Privacy-preserving tabular data synthesis
MarketBench: Evaluating AI Agents as Market ParticipantsLLM agents and tool use → Financial trading and forecasting agents
Evaluation of Prompt Injection Defenses in Large Language ModelsLLM safety and security → Prompt injection and agent security
ZenBrain: A Neuroscience-Inspired 7-Layer Memory Architecture for Autonomous AI SystemsLLM agents and tool use → Agent memory systems
Knowledge Vector of Logical Reasoning in Large Language ModelsLLM reasoning → Reasoning benchmarks and evaluation
Learning Interpretable PDE Representations for Generative Reconstructions with Structured SparsityDiffusion models → Flow matching, time series and science
Latent Inter-Frame Pruning: A Training-Free Method Bridging Traditional Video Compression and Modern Diffusion Transformers for Efficient GenerationDiffusion models → Video generation
Does Machine Unlearning Preserve Clinical Safety? A Risk Analysis for Medical Image ClassificationLLM safety and security → Machine unlearning
ClawTrace: Cost-Aware Tracing for LLM Agent Skill DistillationLLM agents and tool use → Agent skill learning and evolution
Domain-Filtered Knowledge Graphs from Sparse Autoencoder FeaturesMechanistic interpretability → SAE features and their applications
DRACULA: Hunting for the Actions Users Want Deep Research Agents to ExecuteLLM agents and tool use → Web, search, and research agents
ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document ReconstructionLLM reasoning → Multimodal reasoning +1
LegalDrill: Diagnosis-Driven Synthesis for Legal Reasoning in Small Language ModelsLLM reasoning → Reasoning for agents, planning and code
Symmetric Equilibrium Propagation for Thermodynamic Diffusion TrainingDiffusion models → Diffusion foundations and sampling
Domain Fine-Tuning vs. Retrieval-Augmented Generation for Medical Multiple-Choice Question Answering: A Controlled Comparison at the 4B-Parameter ScaleRetrieval-augmented generation → Medical and biomedical RAG
S2G-RAG: Structured Sufficiency and Gap Judging for Iterative Retrieval-Augmented QARetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
ClawMark: A Living-World Benchmark for Multi-Turn, Multi-Day, Multimodal Coworker AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Vision-Language-Action Safety: Threats, Challenges, Evaluations, and MechanismsVision-language-action models → Safety and adversarial attacks
Edit Where You Mean: Region-Aware Adapter Injection for Mask-Free Local Image EditingDiffusion models → Image generation, text-to-image and editing
Agentic Fusion of Large Atomic and Language Models to Accelerate Superconductor DiscoveryLLM agents and tool use → Medical and scientific domain agents
SFT-then-RL Outperforms Mixed-Policy Methods for LLM ReasoningLLM reasoning → Reinforcement learning for reasoning
Transformer as an Euler Discretization of Score-based Variational FlowMixture-of-experts → Expert routing strategies
Prism-Reranker: Beyond Relevance Scoring -- Jointly Producing Contributions and Evidence for Agentic RetrievalRetrieval-augmented generation → Generation and context integration +1
Expert Evaluation of LLM's Open-Ended Legal Reasoning on the Japanese Bar Exam Writing TaskLLM reasoning → Reasoning for agents, planning and code
DynProto: Dynamic Prototype Evolution for Out-of-Distribution DetectionMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Zoom In, Reason Out: Efficient Far-field Anomaly Detection in Expressway Surveillance Videos via Focused VLM Reasoning Guided by Bayesian InferenceLLM reasoning → Chain-of-thought reasoning for driving
HeadRouter: Dynamic Head-Weight Routing for Task-Adaptive Audio Token Pruning in Large Audio Language ModelsEfficient LLM inference and serving → Audio and omni-modal token compression
ZID-Net: Zero-Inference Diffusion Prior Decoupling Network for Single Image DehazingDiffusion models → Inverse problems and image restoration
QuietWalk: Physics-Informed Reinforcement Learning for Ground Reaction Force-Aware Humanoid Locomotion Under Diverse FootwearDeep reinforcement learning methods → RL for robot locomotion and manipulation
FlowPlace: Flow Matching for Chip PlacementDiffusion models → Flow matching, time series and science
BVI-Mamba: Video Enhancement Using a Visual State-Space Model for Low-Light and Underwater EnvironmentsState-space sequence models (S4, Mamba) → Mamba for image restoration
Geometry-Conditioned Diffusion for Occlusion-Robust In-Bed Pose EstimationDiffusion models → 3D, point cloud and scene generation
Structural Enforcement of Goal Integrity in AI Agents via Separation-of-Powers ArchitectureLLM agents and tool use → Agent threats, guardrails and governance
RaV-IDP: A Reconstruction-as-Validation Framework for Faithful Intelligent Document ProcessingRetrieval-augmented generation → Multimodal RAG
Hallo-Live: Real-Time Streaming Joint Audio-Video Avatar Generation with Asynchronous Dual-Stream and Human-Centric Preference DistillationDiffusion models → Audio, speech and music
GraphPlanner: Graph Memory-Augmented Agentic Routing for Multi-Agent LLMsLLM agents and tool use → General LLM agent frameworks and benchmarks
Tandem: Riding Together with Large and Small Language Models for Efficient ReasoningLLM reasoning → Test-time compute and efficient thinking
Move-Then-Operate: Behavioral Phasing for Human-Like Robotic ManipulationVision-language-action models → Failure recovery and robustness
Comparative Study of Weighted and Coupled Second- and Fourth-Order PDEs for Image Despeckling in Grayscale, Color, SAR, and UltrasoundDiffusion models → Medical imaging and segmentation
Tube Diffusion Policy: Reactive Visual-Tactile Policy Learning for Contact-rich ManipulationDiffusion models → Diffusion and flow policies for robotics +1
Thinking Like a Clinician: A Cognitive AI Agent for Clinical Diagnosis via Panoramic Profiling and Adversarial DebateLLM reasoning → Clinical and biomedical applications +1
Talker-T2AV: Joint Talking Audio-Video Generation with Autoregressive Diffusion ModelingDiffusion models → Audio, speech and music
Identity-Decoupled Anonymization for Visual Evidence in Multi-modal Retrieval-Augmented GenerationMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +2
AgentEval: DAG-Structured Step-Level Evaluation for Agentic Workflows with Error Propagation TrackingLLM agents and tool use → Software engineering coding agents
PhysCodeBench: Benchmarking Physics-Aware Symbolic Simulation of 3D Scenes via Self-Corrective Multi-Agent RefinementLLM agents and tool use → Multimodal and embodied agents
CAPSULE: Control-Theoretic Action Perturbations for Safe Uncertainty-Aware Reinforcement LearningDeep reinforcement learning methods → Safe and constrained RL
DLM: Unified Decision Language Models for Offline Multi-Agent Sequential Decision MakingLLM agents and tool use → Reinforcement learning for LLM agents +1
On the Memorization of Consistency Distillation for Diffusion ModelsDiffusion models → Diffusion foundations and sampling
Pref-CTRL: Preference Driven LLM Alignment using Representation EditingLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Oracle Noise: Faster Semantic Spherical Alignment for Interpretable Latent OptimizationDiffusion models → Image generation, text-to-image and editing +1
Z2-Sampling: Zero-Cost Zigzag Trajectories for Semantic Alignment in Diffusion ModelsDiffusion models → Discrete and text diffusion +1
Emotion-Conditioned Short-Horizon Human Pose Forecasting with a Lightweight Predictive World ModelWorld models → Embodied and robot world models
Large Language Model based Interactive Decision-Making for Autonomous DrivingEnd-to-end autonomous driving → Large language models for driving
BurstGP: Enhancing Raw Burst Image Super Resolution with Generative PriorsDiffusion models → Inverse problems and image restoration
Do Prompt-Elicited Trajectories Reflect Training-Time Reward Hacking? A Systematic Study on Monitoring Training-Time Reward Hacking in Code GenerationLLM reasoning → Reinforcement learning for reasoning +1
Agentic Adversarial Rewriting Exposes Architectural Vulnerabilities in Black-Box NLP PipelinesLLM agents and tool use → Agent threats, guardrails and governance
Positioning Generative Artificial Intelligence in STEM Assessment: When to Require, Scaffold, or Restrict Its Use
Flow-by-Flow:Content-Judgment Bypass for Governing AI Output in High-Loss Domains
Proof of Execution: Runtime Verification for Governed AI Agent Actions
Where to Place the Query? Unveiling and Mitigating Positional Bias in In-Context Learning for Diffusion LLMs via Decoding Dynamics
How LLMs Fail and Generalize in RTL Coding for Hardware Design?
Disentangling Linguistic Relatedness from Task Alignment in Cross-Lingual Transfer
Drishti AI-Event Guardian: An Intelligent Real-Time Crowd Monitoring and Emergency Response System for Mass Gathering Events
Crystal Fractional Graph Neural Network for Energy Prediction of High-Entropy Alloys
Path-Based Gradient Boosting for Graph-Level Prediction
Keyword spotting using convolutional neural network for speech recognition in Hindi
Generalization Bounds of Spiking Neural Networks via Rademacher Complexity
Intervention-Based Self-Supervised Learning: A Causal Probe Paradigm for Remote Photoplethysmography
A Coupled Fourth Order Telegraph Diffusion Framework Using Grayscale Indicators for Image Despeckling
Single Image Defogging Using a Fourth-Order Telegraph PDE Guided by Physical Haze Modeling
MAEO: Multiobjective Animorphic Ensemble Optimization for Scalable Large-scale Engineering Applications
A multi-stage soft computing framework for complex disease modelling and decision support: A liver cirrhosis case study
Gromov-Wasserstein Methods for Multi-View Relational Embedding and Clustering
AMAVA: Adaptive Motion-Aware Video-to-Audio Framework for Visually-Impaired Assistance
Machine Learning and Deep Learning Models for Short Term Electricity Price Forecasting in Australia's National Electricity Market
SMSI: System Model Security Inference: Automated Threat Modeling for Cyber-Physical Systems
Integrative neurocybernetic modeling in the era of large-scale neuroscience
LLM-Augmented Traffic Signal Control with LSTM-Based Traffic State Prediction and Safety-Constrained Decision Support
Mammographic Lesion Segmentation with Lightweight Models: A Comparative Study
Geometry Preserving Loss Functions Promote Improved Adaptation of Blackbox Generative Model
MUSIC: Learning Muscle-Driven Dexterous Hand Control
Cardiac Stability Theory: An Axiomatically Grounded Framework for Continuous Cardiac Health Monitoring via Smartphone Photoplethysmography
Risk-Aware Robust Learning: Reducing Clinical Risk under Label Noise in Medical Image Classification
Deep Learning of Solver-Aware Turbulence Closures from Nudged LES Dynamics
Inverting Foundation Models of Brain Function with Simulation-Based Inference
Graph Memory Transformer (GMT)
Empirical Ablation and Ensemble Optimization of a Convolutional Neural Network for CIFAR-10 Classification
Exploring Audio Hallucination in Egocentric Video Understanding
Time-Series Forecasting in Safety-Critical Environments: An Open-Source Package for EU-AI-Act-Compliant Development / Zeitreihenprognose in sicherheitskritischen Umgebungen: Ein Open-Source-Paket f"ur die KI-VO-konforme Entwicklung
Learning Selective LLM Autonomy from Copilot Feedback in Enterprise Customer Support Workflows
Translate or Simplify First: An Analysis of Cross-lingual Text Simplification in English and French
Reheat Nachos for Dinner? Evaluating AI Support for Cross-Cultural Communication of Neologisms
Scalable Production Scheduling: Linear Complexity via Unified Homogeneous Graphs

All 199