Papers

New on arXiv

Sat 1 Aug 2026

229 papers

TopoCompress: Topology Aware Token Compression Algorithm for Distributed Edge MoE InferenceMixture-of-experts → Efficient MoE training and inference systems +1
Cross-Domain Industrial Fault Detection by Causal Mechanism MonitoringState-space sequence models (S4, Mamba) → Time series forecasting with SSMs
Auditing Medical Vision-Language Models on Chest Radiographs: Estimating Reference Agreement Across InstitutionsMultimodal LLMs / vision-language models → Medical vision-language models
SC2-WM: A Self-Correcting World Model with Closed-Loop Feedback for Vision-and-Language Navigation in Continuous EnvironmentsWorld models → World models for navigation
NuclearDiffusion: Text-to-Image Foundation Models for Learning Nuclear Energy ConceptsDiffusion models → Diffusion foundations and sampling
Single Canonical Prompts Underestimate LLM Safety's Surface-Form SensitivityLLM safety and security → Safety alignment, refusal and guardrails
Your Agentic LLMs Secretly Encode Indirect Prompt-Injection Exposure in Hidden StatesLLM agents and tool use → Agent threats, guardrails and governance +1
CADIR: A Cross-Backend Editable Intermediate Representation for Agentic CAD GenerationLLM agents and tool use → Multimodal and embodied agents
Assuming You Knew: Fixing an Epistemic Semantics for Flow Policies Using Agentic AILLM agents and tool use → Agent threats, guardrails and governance
AOSpec: Action and Observation Co-Speculation for Low-Latency Agent ServingLLM agents and tool use → Software engineering coding agents
Bicycle Acrobatics with Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
GeoArbiter: Verifiability-Guided Grounding for Remote-Sensing Multimodal LLMsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Models as Tools: An Agentic Coordination Framework for Unified Multimodal Visual TrackingLLM agents and tool use → Multimodal and embodied agents
AdvPlan-Bench: Adversarial Evaluation of Structured Plan-Generation AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Structure-Aware Semantic Chunking with Title-Chain Prefixes: A 1600-Query Evaluation and the Measurement Trap in Text-Transform AblationsRetrieval-augmented generation → Multimodal RAG
Large language models improve physician accuracy but lead to false relianceRetrieval-augmented generation → Medical and biomedical RAG
OoO-Spec: Out-of-Order Semantic Speculation for Fast Tool CallingLLM agents and tool use → Tool learning and function calling
Turning Interaction History into Execution State: A Runtime Layer for Long-Horizon Coding AgentsLLM agents and tool use → Agent memory systems
AgentSLABench: Evaluating and Benchmarking Agentic Systems Under Resource ConstraintsLLM agents and tool use → General LLM agent frameworks and benchmarks
Measurement Without Validity: The Compounding Reliability Problem in Agentic AI EvaluationLLM agents and tool use → General LLM agent frameworks and benchmarks
DynamicWAM: Dual-Path Motion Conditioning for World-Action Models in Dynamic ManipulationWorld models → Embodied and robot world models
HIERA: Hierarchical Multi-Agent Relevance Assessment for Content Discovery SystemsLLM agents and tool use → Web, search, and research agents
Distill Where You Fail: Recovering Learning Signals of Negative RL-Groups from Adaptive Teacher GuidanceKnowledge distillation → On-policy and self-distillation for LLMs
RAGOCR: Optical Compression of Retrieval-Augmented Text via Visual RepresentationRetrieval-augmented generation → Generation and context integration
FinDeepIndicator: Benchmarking Deep Research Agents in End-to-End Financial Indicator ConstructionLLM agents and tool use → Financial trading and forecasting agents
When Prompts Control Robots: Prompt Injection Attacks in Multi-Agent Robotic SystemsLLM agents and tool use → Agent threats, guardrails and governance +1
LUT: Latent Utility Training for Visual ReasoningLLM reasoning → Multimodal reasoning +1
SelfWAM: A Self-Grounded Unified World Action Model for Fast Robot ControlWorld models → Embodied and robot world models
Adversarial Attacks in Multi-Agent LLM Pipelines: Unveiling Structural Vulnerabilities in Agentic AI ArchitecturesLLM agents and tool use → Agent threats, guardrails and governance
Generated Images Are Easier to Forget: A Machine Unlearning Perspective for Synthetic Image DetectionLLM safety and security → Machine unlearning
Coverage-Driven Adaptive Keyframe Selection for Video UnderstandingMultimodal LLMs / vision-language models → Video understanding with MLLMs
Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent HallucinationsLLM agents and tool use → Medical and scientific domain agents
A Triple-Robustness Analysis of Retrieval-Augmented Generation for Multi-Hop Requirements TraceabilityRetrieval-augmented generation → RAG robustness and adaptive retrieval
OpenART: Scaling Agent Red Teaming via Open-Ended Environment EvolutionLLM safety and security → Prompt injection and agent security
Round-Trip Consistency: Bidirectional Diffusion Models Can Predict Their Own Rollout ErrorsDiffusion models → Flow matching, time series and science
Select-And-Extract: A Lightweight Plugin for Retrieval-Augmented GenerationRetrieval-augmented generation → RAG pipeline design and evaluation
FlowPilot: Real-Time World-Action Modeling for Agile UAV NavigationWorld models → World models for navigation
Escaping Confidence Trap: Evolutionary Decoding for Mathematical Reasoning in Diffusion LLMsLLM reasoning → Diffusion reasoning and test-time scaling
Why Does the Future Branch? Identifiable Closure Tests for Stochastic Physical World ModelsWorld models → Latent predictive world models
InstancePin: Instance-Addressable Layout-to-Image Diffusion via Coordinate PinningDiffusion models → Image generation, text-to-image and editing
Representation Transfer of Foundation Models for Ultra-Widefield Retinal ImagingKnowledge distillation → Distillation for medical imaging
Verification Without Sufficiency: Per-Chunk Filtering Fails on Multi-Hop RAG, and Decomposition Repairs ItRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
Element-Aware Group Learning for E-Commerce Image GenerationLLM reasoning → Reinforcement learning for reasoning
A False Average: Pooled CoT-Monitor Accuracy Conceals a Reasoning-Dependent FragilityLLM reasoning → Prompting and chain-of-thought reasoning
HetRoute Heterogeneous and Cost-aware Collaborative Routing Framework for Distributed Edge MoE InferenceMixture-of-experts → Efficient MoE training and inference systems
Relax Within, Balance Across: Geometry-Guided Load Balancing for Vision-Language Mixture-of-ExpertsMixture-of-experts → Vision mixture-of-experts
TrimMoE A communication aware and adaptive depth framework for distributed edge inferenceMixture-of-experts → Efficient MoE training and inference systems +1
Latency-Tolerant Cloud-Edge Collaborative Vision-Language-Action Models via Emergent Representational SpecializationVision-language-action models → Efficient VLA inference and adaptation
Through the LENS: Local Geometric Decomposition of Vision-Language Model RepresentationsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
DexMani: Human-Derived Manipulability Guidance for Dexterous RotationDeep reinforcement learning methods → RL for robot locomotion and manipulation
DrawAI: Agentic Benchmark and Workflow for Making Raster Images EditableLLM agents and tool use → Multimodal and embodied agents
Disentangling Visuo-Tactile Foresight: Oracle-Guided Interface Discovery for World Action ModelsWorld models → Embodied and robot world models
Robust Watermarks Meet Backdoored Models: Evading Diffusion Semantic Watermarks via Stealthy BackdoorDiffusion models → Robustness, privacy and adversarial aspects +1
Agentic Graph Token ReasoningLLM reasoning → Knowledge- and tool-augmented reasoning +1
DiffuseAgent-MI: Distributionally-Grounded,Tool-Integrated Self-Evolving Agents for Faithful Visual ReasoningLLM reasoning → Multimodal reasoning
Hybrid-Domain Posterior Sampling for Inverse Problems via Latent Flow MatchingDiffusion models → Flow matching, time series and science
Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian LanguagesLLM reasoning → Reasoning benchmarks and evaluation
CURE: Local Uncertainty Repair for Block-Parallel Speculative DecodingEfficient LLM inference and serving → Speculative decoding
Unleashing the Power of Text: Text-Guided Flow Matching for Image Fusion under Complex DegradationsDiffusion models → Image generation, text-to-image and editing
GuideGround: VLM-guided Semantic Understanding and Viewpoint-aware Reasoning for 3D Visual GroundingLLM reasoning → Spatial and embodied reasoning
DreamTraj: Generating 6-DoF Object Trajectories by Reading Unrendered Video Diffusion LatentsDiffusion models → Diffusion and flow policies for robotics
SERL-SQL: Selective Hindsight Distillation for Text-to-SQL Reinforcement Agentic LearningLLM agents and tool use → Reinforcement learning for LLM agents
BANDMAS: Causality-Inspired Semantic Packet Scheduling for Bandwidth-Efficient Multi-Agent CollaborationLLM agents and tool use → General LLM agent frameworks and benchmarks
Reconstruction-Shift Discrimination via Mask-Guided Latent Diffusion for Medical Anomaly DetectionDiffusion models → Medical imaging and segmentation
Distilling Reasoning Traces into Advisory Prompts for Software Engineering TasksLLM reasoning → Reasoning benchmarks and evaluation
Deep Research Pretraining via Predictive NavigationLLM agents and tool use → Web, search, and research agents
MAPLE-Guard: Memory-Aware Link Enforcement Against Memory-Link Poisoning in Multi-Agent SystemsLLM agents and tool use → Agent threats, guardrails and governance
Diagnose Before You Compress: Prediction-Independent Bottleneck Witness Refinement for LLM Serving TracesEfficient LLM inference and serving → Serving systems and request scheduling
TrAC: Trace-Conditioned Answer Consistency for Efficient Uncertainty Quantification in LLMsLLM reasoning → Test-time compute and efficient thinking
SymboUQ: Symbolic Uncertainty Quantification for Spatial Reasoning in LLMsLLM reasoning → Spatial and embodied reasoning
Pretrain on Small Synthetic Data, Scale Large for Free: Symmetry-Aware Foundation Model for Logic Rule InductionSynthetic data and self-improvement for LLMs → Synthetic reasoning data and bootstrapping
The Domestic Unprotected Zone: Algorithmic Governance and the Reproduction of Perpetrator Discourse in Conversational AI
Towards Adaptive Federated Graph Clustering: A Global Community-aware Contrastive Learning-based Approach
Do Existing Preconditioners Improve Biomedical Tabular Foundation Learning? An Empirical Study on TabPFN Optimization
When Does the Best Sampling Temperature Rise with the Budget? Sufficient Conditions for Pass@k
Quantifying Depth Sufficiency in Residual Neural Networks: A First-Order Criterion
In-Context Learning to Assess Built Environment Impacts on Perceived Neighborhood Walkability Among Mobility-impaired Older Adults
Does the Heart Show Your Pain? Tackling the X-ITE Pain Challenge with Self-Supervised ECG Representation Learning
Learning to Adapt Cross-Domain Preferences via Meta-LoRA for LLM Personalization
A Bayes-Markov Neuromorphic Model of Cortical Orientation Selectivity: A Computational Re-implementation and Quantitative Simulation Study
Query Timing Produces Opposite Positional Biases Between LLMs and Humans
The Off-Support Barrier: Why Semantic Safety Constraints Are Not Learning-Problem Invariants, and What Follows for Prior Design, Containment, and Verification
Impact of Dataset Composition on Embedded Real-Time UAV Wildfire Detection Using Compact YOLO Models
P2Voxel: Pyramid Pivot Voxelization for 3D Mesh Tokenization
ZK-SR117: A Chunked Zero-Knowledge Attestation Design for Aggregated Fair-Lending Metrics, with a Control Mapping toward Full SR 11-7 Coverage
CT-HEG: A Bidirectional, Timestamp-Attributed Event Graph for ICU In-Hospital Mortality Prediction - An Architectural Ablation Study
Verifier-Guided Model Discovery for Physical Dynamical Systems with Pretrained Symbolic Transformers
AI Alignment and Fiduciary Obligation
Secure AI Watermarking Framework for IP Protection in Multi-Tenant Cloud Platforms
MBO Scheme for Local Chan--Vese Segmentation
Neuro-Evolved Heuristics for Variable Gapped Common Subsequence Identification
A Sequence-to-Sequence ConvLSTM Approach for Leaf Area Index Forecasting over the South-Central United States
Similarity Weighted Aggregation with Global Differential Privacy for Federated Brain Lesion Segmentation
PhenoStitch: Training-Free Panoptic Crop Mapping from Satellite Image Time Series
Explainable Hybrid Feature Selection for Intrusion Detection in Internet of Medical Things Environments
MIDAL: A Dataset of Math Image Descriptions for Accessible Learning
Kilobyte Models: Neural Networks as a Seed and a Quantized Latent
SparseKAN: Compressing Kolmogorov--Arnold Networks Across Basis Functions, Neurons, and Bits
REIMU: Efficient Heterogeneous Hierarchical Reasoning for SSL-Based Speech Deepfake Detection
Physics-informed neural networks for two-dimensional wall-reactive solute dispersion in canonical shear flows

All 229