Daily Arxiv

New

전 세계에서 발간되는 인공지능 관련 논문을 정리하는 페이지 입니다. 본 페이지는 Google Gemini를 활용해 요약 정리하며, 비영리로 운영 됩니다. 논문에 대한 저작권은 저자 및 해당 기관에 있으며, 요약본 공유 시 출처만 명기하면 됩니다. This service is supported by Google Gemini.

Damage-Aware Bandit Pruning for Vision and Language Transformers

AutoFyn Technical Report: Non-Parametric Expert Iteration for Long-Horizon Agents

When Does Memory Help? A Cost-Aware Evaluation of Long-Term Memory in Tool-Using LLM Agents

CriticGen: Generation-Aware Evaluation as Actionable Feedback

Beyond Right and Wrong: Evaluating Second-order Social Reasoning in Large Language Models

Reducing Hallucinated Transcripts in Whisper via Hallucination Space Projection

IPGeoAI: Transformer-Based Geolocation with LLM Semantic Fusion

From Answers to Interpretations: Rethinking Ambiguity-Induced Aleatoric Uncertainty Estimation in LLMs

Data-Driven Discovery of Composition-Dependent Constitutive Models for Hyperelasticity and Viscoelasticity of Digital Materials

Towards a universal language of concepts: A survey

MaxKernel: Agentic Kernel Generation for TPUs

What Does Multi-Harness RL Learn? Credit Assignment and Portability in Coding Agents

BioSync: Transformer-Based Cross-Modal Fusion for a Multimodal Physiological Digital Biomarker

Rethinking Indirect Prompt Injection as a Test-Time Search Problem

ResLearn-XR: Residual Learning for Network Traffic and Quality-of-Experience-Aware Modeling in Extended Reality

When Quantization Breaks Memory: Recurrent-State Write-Back in Low-Precision Temporal Inference

PerfReasoning: How Well Do LLMs Reason on Hardware Performance?

HarvestBench: Measuring Whether LLM Agents Will Pay to Avoid Killing Animals

Corporate Language Model (CLM): Transforming Tacit and Fragmented Enterprise Knowledge into a Sovereign, Auditable, and Executable Corporate Intelligence Layer

Why Better Models Can Create Riskier Systems: Evidence from LLM Agents in Financial Markets

A Removal Based Approach to Improve LLM Faithfulness at Test-Time

Iris: Climbing to the Search Frontier

Data-Optimized Contingency Screening: A Machine Learning Approach to Power System Security

Harbor Adapters and Harbor-Index: Infrastructure and a Curated Meta-Dataset for Large-Scale Agentic Evaluation

From Matching Models to Recruiting Agents: A Systematized Narrative Review of AI Recruitment Systems, Evaluation, and Governance

EXAONE Forecast for Finance

ViSAR: Training-Free Adaptive-$k$ Retrieval for Visual Document Question Answering

Towards a Foundational Ontology for Identifying and Resolving Contradictions in Dialogue-based Human-Robot Interactions

VoRTeC: Taming Foundation Flow for One-step Real time Video Compression

OmegaUse-SOP: SOP Engineering for Professional Computer Use from Human Demonstrations

Transfer Safety Awareness for Cross-Modal Safety Drift in Multimodal Large Language Models

A Mathematical Theory of Reusable Neural Bases for Network Compression

LatentPress: Context Compression Beyond Text and Vision

Efficiently Estimating Optimal Hyperparameter Scaling Laws through Power-Law Entropy Search

Scientific Agent Skills: A Library of Procedural Knowledge for Research Agents

PAWBench: How Far Are We from Probabilistically Aligned World Modeling?

Safety Does Not Compose: Non-Decaying Loop State for Autonomous LLM Agents

Refusal geometry reflects refusal training: diverse refusal prefixes can raise stable rank and weaken refusal vector ablation attacks

TRACE: A Self-Evolving Skill Bank for Consistent, Limit-Aware LLM Agents

Counterfactual Contrastive Analysis

A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors

WDL-OPD: Weak-Driven On-Policy Distillation via Mixture-Constrained Co-Training

Ask Twice, Look Twice: Prompt Echoing Resolves the Question-First Paradox in Vision-Language Models

PalmClaw: A Native On-Device Agent Framework for Mobile Phones

Learning in Curved Weight Space:Exponential-Linear Weight Reparameterization for Improved Optimization

LLM-Based Test Oracles: Source-of-Authority Taxonomy -- A Systematic Literature Review

KARMA: Knowledge graph-based Automated Reasoning Materialization and Alignment

SNAP-FM: Sparse Nonlinear Accelerated Projection for Physics-Constrained Generative Modeling

Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation

Faithful by Construction: Claim-Anchored Attribution for Multi-Document Summarization

Learning What Not to Forget: Long-Horizon Agent Memory from a Few Kilobytes of Learning

LLMZero: Discovering Adaptive Training Strategies for RL Post-Training via LLM Agents

MeEvo: Metacognitive Evolution Combined with Natural Evolution for Automatic Heuristic Design

Repetition Mismatch: Why Data Mixture Experiments Don't Scale and How to Fix Them

TEVI: Text-Conditioned Editing of Visual Representations via Sparse Autoencoders for Improved Vision-Language Alignment

SV-Detect: AI-generated Text Detection with Steering Vectors

ArcANE: Do Role-Playing Language Agents Stay in Character at the Right Time?

Fixing FOLIO and MALLS: Verified Annotations and an LLM-assisted Framework to Focus Human Relabeling

EntangleCodec: A Unified Discrete Audio Tokenizer via Semantic-Acoustic Entanglement

Argument Collapse: LLMs Flatten Long-Form Public Debate

HARP: Hadamard-Preconditioned Adaptive Rotation Processor for Extreme LLM Quantization

Skill-Conditioned Gated Self-Distillation for LLM Reasoning

EmoDistill: Offline Emotion Skill Distillation for Language Model Agents in Adversarial Negotiation

Identifying AI Web Scrapers Using Canary Tokens

Beyond Reproducibility: Towards Security-Aware Evaluation of Research Artifacts

When Chain-of-Thought Fails, the Solution Hides in the Hidden States

CASCADE: A Component Ablation and Corpus Audit of a Layered Local Defense for MCP-Based Systems

LLM Evaluation as Tensor Completion: Low Rank Structure and Semiparametric Efficiency

One Model to Translate Them All? A Journey to Mount Doom for Multilingual Model Merging

LRConv-NeRV: Low Rank Convolution for Efficient Neural Video Compression

The Landscape of Generative AI in Information Systems: A Synthesis of Secondary Reviews and Research Agendas

PeroMAS: A Multi-agent System of Perovskite Material Discovery

FedPS: Federated Preprocessing for structured data via aggregated Statistics

Ex-Omni: Enabling 3D Facial Animation Generation for Omni-modal Large Language Models

F-GRPO: Don't Let Your Policy Learn the Obvious and Forget the Rare

Temperature Scaling Attack Disrupting Model Confidence in Federated Learning

VoxPrivacy: A Benchmark for Evaluating Interactional Privacy of Speech Language Models

Relational Linearity is a Predictor of Hallucinations

HOMURA: Taming the Sand-Glass for Time-Constrained LLM Translation via Reinforcement Learning

Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models

FADTI: Fourier and Attention Driven Diffusion for Multivariate Time Series Imputation

Evolving Excellence: Automated Optimization of LLM-based Agents

Mixed Data Clustering Survey and Challenges

AnyBox: Efficient Zero-Shot 9DoF Pose Estimation of Boxes for Robotic Manipulation

Short-Window Sliding Learning for Real-Time Violence Detection via LLM-based Auto-Labeling

User Perceptions vs. Proxy LLM Judges: Privacy and Helpfulness in LLM Responses to Privacy-Sensitive Scenarios

EasySteer: A Unified Framework for High-Performance and Extensible LLM Steering

Human Psychometric Questionnaires Mischaracterize LLM Behavior

Decentralized Vision-Based Autonomous Aerial Wildlife Monitoring

Measuring Harmfulness of Computer-Using Agents

Medical Reasoning in the Era of LLMs: A Systematic Review of Enhancement Techniques and Applications

ScoreMix: Synthetic Data Generation by Score Composition in Diffusion Models Improves Recognition

LightEMMA: A Longitudinal Evaluation of Vision-Language Models for Autonomous Driving

Sionna RT: Technical Report

AgentRM: Enhancing Agent Generalization with Reward Modeling

LDC: Learning to Generate Research Idea with Dynamic Control

Data Market Design through Deep Learning

From Analytics to Tumor Boards: An Evidence-Linked Multi-Agent Workflow for Oncology Feature Extraction

VideoHarness-RSI: Recursive Harness Self-Improvement for Long-Video Understanding with Frozen Vision-Language Models

AI Agents Push Humans Out of the Loop

Made with Slashpage