KOINEU

Beyond Single-Agent Safety: A Taxonomy of Risks in LLM-to-LLM Interactions

1. 연구 배경 및 문제 제기 단일‑에이전트 안전 패러다임의 한계 기존 안전 기술(RLHF, 프롬프트 엔지니어링, 출력 모더레이션 등)은 점별 (pointwise) 제어에 초점을 맞춘다. 이는 “하나의 모델 ↔ 하나의 사용자”라는 이원적(dyadic) 상황을 전제로 하며, 모델의 출력이 외부 시스템에 재투입되는 경우를 고려하지 않는다. LLM‑to‑LLM 생태계의 급성장 AutoGen, CAMEL, SWE‑agent, Voyager 등에서 보듯, LLM이 도구, 메모리, 다른 LLM과 연계되는 멀티‑에이전트 구조가 실무와 연구 모두

Beyond Single-Agent Safety: A Taxonomy of Risks in LLM-to-LLM Interactions

Dynamical modeling of nonlinear latent factors in multiscale neural activity with real-time inference

Exploration vs. Fixation: Scaffolding Divergent and Convergent Thinking for Human-AI Co-Creation with Generative Models

LoopBench: Discovering Emergent Symmetry Breaking Strategies with LLM Swarms

A DeepSeek-Powered AI System for Automated Chest Radiograph Interpretation in Clinical Practice

Statistical Arbitrage in Polish Equities Market Using Deep Learning Techniques

AI-Driven Expansion and Application of the Alexandria Database

Towards Efficient Hypergraph and Multi-LLM Agent Recommender Systems

입력 크기와 무관한 시각 인코더 MambaEye

AgriRegion: Region-Aware Retrieval for High-Fidelity Agricultural Advice

AutoICE: Automatically Synthesizing Verifiable C Code via LLM-driven Evolution

Automated Risk-of-Bias Assessment of Randomized Controlled Trials: A First Look at a GEPA-trained Programmatic Prompting Framework

ReactorFold: Generative discovery of nuclear reactor cores via emergent physical reasoning

An Optimal Policy for Learning Controllable Dynamics by Exploration

Remoe: Towards Efficient and Low-Cost MoE Inference in Serverless Computing

The Wisdom of Deliberating AI Crowds: Does Deliberation Improve LLM-Based Forecasting?

Distill, Forget, Repeat: A Framework for Continual Unlearning in Text-to-Image Diffusion Models

Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images

Safe Path Planning and Observation Quality Enhancement Strategy for Unmanned Aerial Vehicles in Water Quality Monitoring Tasks

Scalable Decision Focused Learning via Online Trainable Surrogates

Software Vulnerability Management in the Era of Artificial Intelligence: An Industry Perspective

Unavoidable patterns and plane paths in dense topological graphs

CORE: Concept-Oriented Reinforcement for Bridging the Definition-Application Gap in Mathematical Reasoning

Deconstructing Generative Diversity: An Information Bottleneck Analysis of Discrete Latent Generative Models

DUET: Agentic Design Understanding via Experimentation and Testing

Attention in Motion: Secure Platooning via Transformer-based Misbehavior Detection

Cognitive Control Architecture (CCA): A Lifecycle Supervision Framework for Robustly Aligned AI Agents

Multi-Intent Spoken Language Understanding: Methods, Trends, and Challenges

PHANTOM: Progressive High-fidelity Adversarial Network for Threat Object Modeling

Visual Funnel: Resolving Contextual Blindness in Multimodal Large Language Models

Improving Local Fidelity Through Sampling and Modeling Nonlinearity

Reusability in MLOps: Leveraging Ports and Adapters to Build a Microservices Architecture for the Maritime Domain

Assignment-Routing Optimization: Solvers for Problems Under Constraints

Earth radius from a single sunrise image: a classroom-ready activity

Memory as Resonance: A Biomimetic Architecture for Infinite Context Memory on Ergodic Phonetic Manifolds

Tractatus Quanticum

AdaptVision: Efficient Vision-Language Models via Adaptive Visual Acquisition

JMMMU-Pro: Image-based Japanese Multi-discipline Multimodal Understanding Benchmark via Vibe Benchmark Construction

Language Models as Semantic Teachers: Post-Training Alignment for Medical Audio Understanding

PyBangla at BLP-2025 Task 2: Enhancing Bangla-to-Python Code Generation with Iterative Self-Correction and Multilingual Agents

Audited Skill-Graph Self-Improvement for Agentic LLMs via Verifiable Rewards, Experience Synthesis, and Continual Memory

CODE ACROSTIC: Robust Watermarking for Code Generation

LabelFusion: Learning to Fuse LLMs and Transformer Classifiers for Robust Text Classification

Information-Dense Reasoning for Efficient and Auditable Security Alert Triage

SIMA 2: A Generalist Embodied Agent for Virtual Worlds

AI로 보는 저비용 체지방률 추정 이미지와 인체계측 데이터 활용

ForCM: Forest Cover Mapping from Multispectral Sentinel-2 Image by Integrating Deep Learning with Object-Based Image Analysis

PCIA: A Path Construction Imitation Algorithm for Global Optimization

Post-Cold War Diaspora of Russian Particle Physicists

경량 에이전트 코어 Xmodel‑2.5: µP 기반 파라미터 전이와 FP8 혼합 정밀도 학습

< Category Statistics (Total: 5055) >

Start searching

No results found