Daily Arxiv

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, 공유 시 출처만 명기하면 됩니다.

LawThinker: A Deep Research Legal Agent in Dynamic Environments

Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments

MEME: Modeling the Evolutionary Modes of Financial Markets

AlphaPROBE: Alpha Mining via Principled Retrieval and On-graph biased evolution

When Should LLMs Be Less Specific? Selective Abstraction for Reliable Long-Form Text Generation

From Atoms to Trees: Building a Structured Feature Forest with Hierarchical Sparse Autoencoders

Intelligent AI Delegation

Talk2DM: Enabling Natural Language Querying and Commonsense Reasoning for Vehicle-Road-Cloud Integrated Dynamic Maps with Large Language Models

Prototype Transformer: Towards Language Model Architectures Interpretable by Design

Revis: Sparse Latent Steering to Mitigate Object Hallucination in Large Vision-Language Models

Predicting LLM Output Length via Entropy-Guided Representations

PuYun-LDM: A Latent Diffusion Model for High-Resolution Ensemble Weather Forecasts

Hi-SAM: A Hierarchical Structure-Aware Multi-modal Framework for Large-Scale Recommendation

Detecting RLVR Training Data via Structural Convergence of Reasoning

Beyond End-to-End Video Models: An LLM-Based Multi-Agent System for Educational Video Generation

FlowMind: Execute-Summarize for Structured Workflow Generation from LLM Reasoning

RELATE: A Reinforcement Learning-Enhanced LLM Framework for Advertising Text Generation

How to Optimize Multispecies Set Predictions in Presence-Absence Modeling ?

TSR: Trajectory-Search Rollouts for Multi-Turn RL of LLM Agents

AIR: Improving Agent Safety through Incident Response

Text2GQL-Bench: A Text to Graph Query Language Benchmark [Experiment, Analysis & Benchmark]

Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMs

Beyond Parameter Arithmetic: Sparse Complementary Fusion for Distribution-Aware Model Merging

ThinkRouter: Efficient Reasoning via Routing Thinking between Latent and Discrete Spaces

Beyond Pixels: Vector-to-Graph Transformation for Reliable Schematic Auditing

Right for the Wrong Reasons: Epistemic Regret Minimization for Causal Rung Collapse in LLMs

Benchmark Health Index: A Systematic Framework for Benchmarking the Benchmarks of LLMs

PhyNiKCE: A Neurosymbolic Agentic Framework for Autonomous Computational Fluid Dynamics

Quark Medical Alignment: A Holistic Multi-Dimensional Alignment and Collaborative Optimization Paradigm

Do MLLMs Really Understand Space? A Mathematical Reasoning Evaluation

Neuro-Symbolic Multitasking: A Unified Framework for Discovering Generalizable Solutions to PDE Families

When Agents Disagree With Themselves: Measuring Behavioral Consistency in LLM-Based Agents

scPilot: Large Language Model Reasoning Toward Automated Single-Cell Analysis and Discovery

MAPLE: Modality-Aware Post-training and Learning Ecosystem

The Five Ws of Multi-Agent Communication: Who Talks to Whom, When, What, and Why -- A Survey from MARL to Emergent Language and LLMs

Learning to Configure Agentic AI Systems

SemaPop: Semantic-Persona Conditioned Population Synthesis

Budget-Constrained Agentic Large Language Models: Intention-Based Planning for Costly Tool Use

CausalAgent: A Conversational Multi-Agent System for End-to-End Causal Inference

Human-Inspired Continuous Learning of Internal Reasoning Processes: Learning How to Think for Adaptive AI Systems

AgentLeak: A Full-Stack Benchmark for Privacy Leakage in Multi-Agent LLM Systems

Credit Where It is Due: Cross-Modality Connectivity Drives Precise Reinforcement Learning for MLLM Reasoning

Distributionally Robust Cooperative Multi-Agent Reinforcement Learning via Robust Value Factorization

TRACER: Trajectory Risk Aggregation for Critical Episodes in Agentic Reasoning

Causal-JEPA: Learning World Models through Object-Level Latent Interventions

ReplicatorBench: Benchmarking LLM Agents for Replicability in Social and Behavioral Sciences

Pushing Forward Pareto Frontiers of Proactive Agents with Behavioral Agentic Optimization

AgentNoiseBench: Benchmarking Robustness of Tool-Using LLM Agents Under Noisy Condition

Bi-Level Prompt Optimization for Multimodal LLM-as-a-Judge

Dissecting Subjectivity and the "Ground Truth" Illusion in Data Annotation

Voxtral Realtime

On Decision-Valued Maps and Representational Dependence

Latent Generative Solvers for Generalizable Long-Term Physics Simulation

Explaining AI Without Code: A User Study on Explainable AI

Time-to-Event Transformer to Capture Timing Attention of Events in EHR Time Series

ENIGMA: EEG-to-Image in 15 Minutes Using Less Than 1% of the Parameters

Learning Self-Interpretation from Interpretability Artifacts: Training Lightweight Adapters on Vector-Label Pairs

Are More Tokens Rational? Inference-Time Scaling in Language Models as Adaptive Resource Rationality

Confounding Robust Continuous Control via Automatic Reward Shaping

ECHO: An Open Research Platform for Evaluation of Chat, Human Behavior, and Outcomes

ERGO: Excess-Risk-Guided Optimization for High-Fidelity Monocular 3D Gaussian Splatting

From Classical to Topological Neural Networks Under Uncertainty

The Complexity of Bayesian Network Learning: Revisiting the Superstructure

KORAL: Knowledge Graph Guided LLM Reasoning for SSD Operational Analysis

Transforming Policy-Car Swerving for Mitigating Stop-and-Go Traffic Waves: A Practice-Oriented Jam-Absorption Driving Strategy

ImprovEvolve: Ask AlphaEvolve to Improve the Input Solution and Then Improvise

Blockwise Advantage Estimation for Multi-Objective RL with Verifiable Rewards

Self-Evolving Recommendation System: End-To-End Autonomous Model Optimization With LLM Agents

Quantum Integrated Sensing and Computation with Indefinite Causal Order

Internalizing Meta-Experience into Memory for Guided Reinforcement Learning in Large Language Models

Versor: A Geometric Sequence Architecture

When the Prompt Becomes Visual: Vision-Centric Jailbreak Attacks for Large Image Editing Models

Towards Autonomous Mathematics Research

Cosmo3DFlow: Wavelet Flow Matching for Spatial-to-Spectral Compression in Reconstructing the Early Universe

EvoCodeBench: A Human-Performance Benchmark for Self-Evolving LLM-Driven Coding Systems

EVA: Towards a universal model of the immune system

Anatomy-Preserving Latent Diffusion for Generation of Brain Segmentation Masks with Ischemic Infarct

Beyond SMILES: Evaluating Agentic Systems for Drug Discovery

Omni-Safety under Cross-Modality Conflict: Vulnerabilities, Dynamics Mechanisms and Efficient Alignment

AD$^2$: Analysis and Detection of Adversarial Threats in Visual Perception for End-to-End Autonomous Driving Systems

NMRTrans: Structure Elucidation from Experimental NMR Spectra via Set Transformers

MalMoE: Mixture-of-Experts Enhanced Encrypted Malicious Traffic Detection Under Graph Drift

PRISM-XR: Empowering Privacy-Aware XR Collaboration with Multimodal Large Language Models

PEST: Physics-Enhanced Swin Transformer for 3D Turbulence Simulation

Exploring Semantic Labeling Strategies for Third-Party Cybersecurity Risk Assessment Questionnaires

Red-teaming the Multimodal Reasoning: Jailbreaking Vision-Language Models via Cross-modal Entanglement Attacks

On the Use of a Large Language Model to Support the Conduction of a Systematic Mapping Study: A Brief Report from a Practitioner's View

Silence Routing: When Not Speaking Improves Collective Judgment

When LLMs get significantly worse: A statistical approach to detect model degradations

Can Large Language Models Implement Agent-Based Models? An ODD-based Replication Study

Anonymization-Enhanced Privacy Protection for Mobile GUI Agents: Available but Invisible

Multimodal Information Fusion for Chart Understanding: A Survey of MLLMs -- Evolution, Limitations, and Cognitive Enhancement

Multi-encoder ConvNeXt Network with Smooth Attentional Feature Fusion for Multispectral Semantic Segmentation

Reverse-Engineering Model Editing on Language Models

AgentTrace: A Structured Logging Framework for Agent System Observability

TokaMark: A Comprehensive Benchmark for MAST Tokamak Plasma Models

The Anatomy of the Moltbook Social Graph

"Humans welcome to observe": A First Look at the Agent Social Network Moltbook

A Practical Guide to Agentic AI Transition in Organizations

Large Language Models Predict Functional Outcomes after Acute Ischemic Stroke

Wiki-TabNER: Integrating Named Entity Recognition into Wikipedia Tables

Created by

Haebom

Category

Empty

저자

Aneta Koleva, Martin Ringsquandl, Ahmed Hatem, Thomas Runkler, Volker Tresp

개요

본 논문은 기존 표 해석 작업용 데이터셋의 단순화 문제를 지적하며, 보다 현실적인 복잡한 표를 포함하는 새로운 데이터셋 Wiki-TabNER를 제시합니다. Wiki-TabNER는 셀당 여러 개체를 포함하고 DBpedia 클래스를 사용하여 명명된 개체를 라벨링한 것이 특징입니다. 표 내부의 명명된 개체 인식(NER) 작업을 위한 설계이지만, 개체 연결 작업 평가에도 활용 가능합니다. 본 논문에서는 Wiki-TabNER 데이터셋의 특징과 라벨링 과정, 그리고 대규모 언어 모델 평가를 위한 프롬프팅 프레임워크를 설명하고, 모델의 어려움과 데이터셋의 한계를 분석하는 질적 분석 결과를 제시합니다.

시사점, 한계점

•

시사점:

◦

현실적인 복잡성을 반영하는 새로운 표 데이터셋 Wiki-TabNER 제공

◦

표 내 NER 작업 평가를 위한 새로운 기준 마련

◦

대규모 언어 모델의 표 해석 성능 평가를 위한 프롬프팅 프레임워크 제시

◦

표 해석 모델의 한계점과 데이터셋 개선 방향에 대한 통찰 제공

•

한계점:

◦

Wiki-TabNER 데이터셋의 규모 및 다양성 제한 (향후 확장 필요성)

◦

제시된 프롬프팅 프레임워크의 일반화 가능성 제한 (특정 모델/과제에 대한 최적화 가능성)

◦

질적 분석의 주관성 (객관적인 평가 지표 추가 필요)

Made with Slashpage