Papers

New on arXiv

Sat 25 Jul 2026

180 papers

Silence Is Endorsement: Verification-Status Laundering in LLM Agent PipelinesLLM agents and tool use → Agent threats, guardrails and governance
Harnessing agent memory to build lifelong AI partners for materials scientistsLLM agents and tool use → Medical and scientific domain agents
Hierarchical Copula-Gumbel-Top-K Routing: Two-Sided Dependence Control for Frozen Mixture-of-Experts at Fixed Per-Token Routing LawsMixture-of-experts → Expert routing strategies
LARA: Lightweight Adapters in the Residual Stream for Composable Adaptation and AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
SearchArt: Training Long-Horizon Search Agent with Scalable Synthetic and Verified TaskLLM agents and tool use → Web, search, and research agents
UltraViT: Latency-Optimized On-device Vision Encoder for Large Vision-Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
On the Impossibility of Unbiased and Length-Invariant Policy Optimization with Outcome RewardsLLM reasoning → Reinforcement learning for reasoning
Logit-Coordinate Generative Models for Mixed Continuous-Categorical Tabular DataDiffusion models → Flow matching, time series and science
SPRKD: Effective Knowledge Distillation for Deep Neural Networks via Saddle Region ApproximationKnowledge distillation → General knowledge distillation methods
The Gate Always Closes: On Injecting Auxiliary Signals into Frozen Vision-Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
AllocBench: Measuring Online Tool Allocation Capability in LLM AgentsLLM agents and tool use → Tool learning and function calling
Context-Adaptive Inference: A Unified Statistical and Foundation-Model ViewMixture-of-experts → Medical and biomedical MoE
RareLens: Towards End-to-End Rare Disease Care via Aligning Divergent Large Language Model ReasoningLLM reasoning → Clinical and biomedical applications
Co-Evolving Graph and Text Memory for Training-Free Multi-Hop Question AnsweringLLM reasoning → Knowledge- and tool-augmented reasoning
What CLIP Knows but Cannot Say: Recovering Negation from Frozen Intermediate FeaturesMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
WaveZip: Wavelet-Driven Space-Time Decoupling for Video Token CondensationMultimodal LLMs / vision-language models → Visual token pruning and efficiency
SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use AgentsLLM agents and tool use → GUI and mobile agents +1
From Score Learning to Discretized Sampling: An End-to-End Generalization Analysis of Diffusion ModelsDiffusion models → Diffusion foundations and sampling
OmniScope: Modality-Decoupled Token Compression for Omnimodal Large Language ModelsEfficient LLM inference and serving → Audio and omni-modal token compression
Fashion-3DLR: A Controllable 3D Garment Generation Using Pairwise Fashion Elements for Intelligent DesignDiffusion models → 3D, point cloud and scene generation
Towards Dual-Brain Minimal Sufficient Representation for Vision-Language NavigationWorld models → World models for navigation
CachedSearch: Training-Free Cached Exploration for Test-Time Search in Video DiffusionDiffusion models → Video generation
In-Context Learning as Implicit Policy GradientDeep reinforcement learning methods → RL training of language and agent models
False Prophets: On the Security of World Models in Agentic SystemsWorld models → World model security and GUI agents +1
Self-Boosting Vision-Language Models with Noisy Student On-Policy Self-DistillationMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
AgentOmnia: Scaling Agentic Models for Full-Scenario ApplicationsLLM agents and tool use → General LLM agent frameworks and benchmarks
SQBench: A Benchmark for Evaluating Task Delivery by Language-Model Agents in Production-Oriented WorkflowsLLM reasoning → Reasoning for agents, planning and code +1
The Curse of Precision: A Data Scaling Law for High-Precision Robotic ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
Decoding the Skew: Distribution-Aware MoE Inference with Adaptive Kernel DispatchMixture-of-experts → Efficient MoE training and inference systems
LoRA for Gender-Inclusive Rewriting and Activation Steering for Counter-Narrative GenerationMechanistic interpretability → Activation steering methods
Traceable LLM Reasoning for Fake-Order Fraud DetectionLLM reasoning → Reasoning for agents, planning and code
DishSeg24k: A Large-Scale Benchmark for Food Segmentation with Stochastic Expert DecodingMixture-of-experts → Vision mixture-of-experts
ADAGE: A Language-Agnostic Pipeline for Analogical Reasoning EvaluationLLM reasoning → Reasoning benchmarks and evaluation
SymStep: Symbolic Step Verification for Logical ReasoningLLM reasoning → Reasoning benchmarks and evaluation
Through the Bottleneck: How Multi-head Latent Attention Separates Content from Position in Language ModelsEfficient LLM inference and serving → KV cache optimization
Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Structured Redundancy Modeling for Efficient Visual Token Pruning in High-Resolution MLLMsMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
Stress-testing large language model agents in a robotic chemistry laboratoryLLM agents and tool use → Web, search, and research agents
Online Policy Evaluation for MDPs with Dynamic UBSR MeasuresDeep reinforcement learning methods → RL theory and sample complexity
All in One: Generative Modeling as Mean-Field Game DesignDiffusion models → Diffusion and flow policies for robotics
OmniMate: Open-Ended Real-Time Streaming Audio-Visual Generation for Interactive AvatarsDiffusion models → Audio, speech and music
Reason Popper-ly: Patching In-Context Reasoning with Inductive Logic ProgrammingLLM reasoning → Prompting and chain-of-thought reasoning
Mask2Shield: Strengthening LLM Safety against Neuron-Pruning AttacksLLM safety and security → Safety alignment, refusal and guardrails
VecTree-RAG: An Agentic Retrieval-Augmented Generation Framework Combining Vector and Tree Retrieval for Efficiency and AccuracyRetrieval-augmented generation → Agentic RAG, agents and memory +1
Real2Sim2Real for Vision-Language-Action Manipulation: An AMD ROCm-Based PipelineVision-language-action models → Generalist VLA models and benchmarks
Label-free Industrial Fault Detection via Adversarial Inverse Reinforcement Learning: A System for Run-to-Failure PrognosticsOffline RL and imitation learning from data → Inverse reinforcement learning +1
Finite-Time Analysis of the Natural Policy Gradient in Finite-Horizon Markov Decision ProcessesDeep reinforcement learning methods → RL theory and sample complexity
PriSAR: 3D Geometric-Prior-Guided Diffusion for Parameter-Controlled SAR Image GenerationDiffusion models → Inverse problems and image restoration
ConsistencyGate: Preventing Memory Contamination in LLM Agents via Self-Consistency Admission ControlLLM reasoning → Reasoning for agents, planning and code +1
When the Preconditioning Exponent Turns Negative: Learning-Rate Coupling and Cross-Environment Generalization
Towards Secure Cloud-Native Computing: Unveiling Kubernetes Misconfigurations with Large Language Models
Scene-Conditioned Relation Routing for urban cellular activity forecasting
Foundations of Stochastic Lexical Calculus: Semantic Descent and Random Dynamics on Probability Simplices
JointMatch: A Unified Heterogeneous Graph Neural Solver for Large-Scale Ride-Sharing Matching
Towards a Unified Modality-Agnostic Multimodal Framework for Cognitive Workload Assessment
Adaptive Interpolatory Curve Subdivision with Learned Local Angles
A Heisenberg Lift Descriptor for Order Sensitive Online Handwriting Recognition
CTAN: Cycle-Temporal Attention Network for Embodied Audio-Visual Navigation
A derivative-fidelity failure mode in physics-informed neural networks: strengthened benchmark evidence from function-value training
Identifiability-Aware Source Apportionment in City-Scale Advection-Diffusion Systems
Not All EEG Moments Are Equal: Position-Adaptive Time Scheduling for EEG Generation
Fracture Risk Prediction in Adults Over 50 Years Old Using DXA and EHR: Comparison of Traditional and Machine Learning Models in Two Large Cohorts
Guarantees on Dynamical System Distinguishability for LLM Token Generation
The Checking Problem: What must be true before AI ships in a regulated firm
Sensitivity Analysis of GRU, LSTM and Transformer Encoder in Classification of Automated Driving Systems
Hold-Out Self-Validation Cannot Certify Photogrammetric Accuracy: Saturation and Blindness to Coherent Distortion
Beyond Predictive Accuracy: A Reliability-Aware Audit of Molecular Representations for Human Olfaction
Extremal Chowla sets and their linear analogues: A human-AI mathematical investigation using Co-Scientist
Music-Source-Separation-Training (MSST): A Unified Framework for Training and Evaluating Music Demixing Models
Inference-Time Consensus for Mitigating Hidden Behaviors from LLM Fine-Tuning
Key-Interval A*: Accelerating Grid Pathfinding via Structural Abstraction
When Can Depth Replace Precision? A Resource Theory of Quantized Neural Computation
Rendering on Real Silicon: GPU Render-Timing as a Passive, AI-Resistant CAPTCHA Signal
Constraint-Aware Influence Estimation in Deep Constrained Learning
Confidently Wrong: Exception Chain Collapse in Frontier LLM Rule Evaluation
Semantic Semi-Incremental Data-Association-Free Object SLAM
When Activation Oracles Learn Not to Read: Concept-Specific Blind Spots in Fine-Tuned Oracles
Predict before you train: Scaling Laws for particle physics foundation models
Investigating the Visual Cues of CNNs for Vascular Segmentation: A Case Study in Microscopy and Fundus Imaging
Bitcoin Price Direction Prediction via Regime-Aware Multi-Modal Fusion of Social Sentiment and Technical Features
Explaining BiomedCLIP with Weighted Banzhaf Interactions Supported by Tree-Gram Parsing
Fair Division with Strictly Increasing Valuations: A Tight Threshold for Two-Agent EF1 and PO
Codebook Capacity Governs Perceptual Quality Across Resolutions in Hierarchical Discrete Video Compression
On AI Safety and Security Technical Debt in Engineering AI-Enabled Systems
Joint Optimization for Greedy Longest-match Tokenization
Hallucination Rates in Language Generation
Exploration of the generative capabilities of Boltzmann machines applied to social systems under the majority rule
Trainable Nonexpansive Denoisers for Contractive Image Reconstruction
BERT-based Models vs. Large Language Models for Low-Resource Named Entity Recognition: A Comparative Study on Marathi
Patient-Agnostic Synthetic Pretraining for Efficient Patient-Specific Intraoperative 2D/3D Registration
Stabilizing Deep Reconstruction Operators with Contractive Anchoring
Does Graph Compression Preserve Signal Propagation?
Neural operator discovery from heterogeneous trajectories
Training with (Swap) Regret Loss in a Single-Layer Self-Attention Model: A Case Study on the Probability Simplex
ESF-Bench: Benchmarking Challenging Slot-Filling Scenarios for Real-World Enterprise Applications
IKS-Instruct: A 24,000-Example Multilingual Dataset for Teaching Language Models Indian Knowledge Systems
A Reference-Free Framework for Evaluating Single-Frame ISP Pipelines
BHARATI: Morphology-Aware Tokenizers for Classical Indian Languages with Subword Fertility Analysis
Ordered Network Analysis of Epistemic Emotions during Collaborative Problem Solving
Online Fair Division with Budget Constraints

All 180