TechPulseTelegram
← PrevThu, Aug 13Next β†’

πŸ€– AI Agent19 items

Latest AI agent research papers, analysis, and improvement insights

semantic_scholarresearch

EMAS: Stabilizing Multi-Agent System Evolution through Evidence-Guided Revision

κ²½ν—˜μœΌλ‘œλΆ€ν„° MAS ν† ν΄λ‘œμ§€μ™€ ν”„λ‘¬ν”„νŠΈλ₯Ό μžλ™ κ°œμ„ ν•˜λŠ” ν”„λ ˆμž„μ›Œν¬

semantic_scholarresearch

MoRSE: Task-Oriented Multi-Agent System with Mixture of Role-Subtask Experts

LLM 기반 닀쀑 μ—μ΄μ „νŠΈμ— μ—­ν•  νŠΉν™”μ™€ λ§€κ°œλ³€μˆ˜ 적응 μΆ”κ°€

blog_openairesearch

From assistance to execution: How enterprises put AI to work

κΈ°μ—…μ˜ AI μ—μ΄μ „νŠΈ λ„μž… 및 ν™œμš© λ°©μ‹μ˜ μ‹€μ œ 사둀 연ꡬ

blog_huggingfaceresearch

Meta is back with Muse Glimmer: local, agentic, multimodal, and open source

λ‘œμ»¬μ—μ„œ μ‹€ν–‰λ˜λŠ” μ˜€ν”ˆμ†ŒμŠ€ 에이전틱 λ©€ν‹°λͺ¨λ‹¬ κΈ°λ³Έ λͺ¨λΈ μ†Œκ°œ

hackernewsresearch

Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows

둜컬 μ—μ΄μ „νŠΈ μ›Œν¬ν”Œλ‘œμš° μ΅œμ ν™”λœ κ²½λŸ‰ κΈ°λ³Έ λͺ¨λΈμ˜ 상세 기술

hackernewsresearch

OpenChamber: An Agentic Development Environment

AI μ—μ΄μ „νŠΈ 개발 및 디버깅을 μœ„ν•œ 톡합 개발 ν™˜κ²½

semantic_scholarresearch

Advancing MLLM-based UAV Image Understanding and Reasoning: A Benchmark and a Training-Free Multi-Agent System

무인항곡기 μ˜μƒ 이해λ₯Ό μœ„ν•œ 닀쀑 μ—μ΄μ „νŠΈ μ‹œμŠ€ν…œκ³Ό 톡합 벀치마크 μ œμ‹œ

semantic_scholarresearch

Who Are You Explaining To? A Multi-Agent System for Audience-Aware XAI Narratives

λ‹€μ–‘ν•œ 청쀑을 μœ„ν•΄ AI μ„€λͺ…을 λ§žμΆ€ν˜•μœΌλ‘œ μƒμ„±ν•˜λŠ” MAS

semantic_scholarresearch

An on-premises end-to-end automated forecasting multi-agent system for the energy domain

μ—λ„ˆμ§€ λΆ„μ•Όμ—μ„œ μ˜ˆμΈ‘μ„ μžλ™ν™”ν•˜λŠ” μ—”λ“œνˆ¬μ—”λ“œ MAS

semantic_scholarresearch

AI4S Telescope: a multi-agent embodied intelligence system for hypothesis-driven astronomical observation

자율적 우주 κ΄€μΈ‘ μˆ˜ν–‰ν•˜λŠ” 닀쀑 μ—μ΄μ „νŠΈ 지λŠ₯ μ‹œμŠ€ν…œ

blog_openairesearch

Virgin Atlantic sharpens customer journeys with ChatGPT Work

ChatGPT Workλ₯Ό ν™œμš©ν•œ κΈ°μ—… AI μ—μ΄μ „νŠΈμ˜ μ‹€μ œ 운영 사둀

hackernewsresearch

Qwen3.8 Max now ranked as the best overall model by agentic index

Qwen3.8 Max의 μ—μ΄μ „νŠΈ μ„±λŠ₯ μš°μˆ˜μ„± 벀치마크 κ²°κ³Ό

hackernewsresearch

Show HN: Needle2: 14MB agentic LLM for phones, wearables, smart home and robots

엣지 λ””λ°”μ΄μŠ€ 싀행을 μœ„ν•œ μ΄ˆκ²½λŸ‰ 에이전틱 LLM λͺ¨λΈ

hackernewsresearch

Mythos Attempted to Social Engineer Open Source Maintainer to Merge Malware

AI μ—μ΄μ „νŠΈμ˜ μ•…μ˜μ  μ‚¬μš©μœΌλ‘œ μΈν•œ λ³΄μ•ˆ μœ„ν˜‘ 사둀 뢄석

semantic_scholarresearch

POSTER: A Communication-Centric Multi-Agent System for Persistent UAV Formation Control

자율 리더십 μ „ν™˜μœΌλ‘œ κ΅¬ν˜„ν•œ λΆ„μ‚° UAV νŽΈλŒ€ μ œμ–΄ μ‹œμŠ€ν…œ

hackernewsresearch

WorldClaw Agentic 3D open-world generation at scale

λŒ€κ·œλͺ¨ 3D ν™˜κ²½μ„ μžλ™μƒμ„±ν•˜λŠ” 에이전틱 AI μ‹œμŠ€ν…œ

hackernewsresearch

Human vs. AI – Diff-based line-level provenance for text under agentic editing

AI μ—μ΄μ „νŠΈ νŽΈμ§‘μ˜ 투λͺ…성을 μœ„ν•œ 변경사항 좔적 기술

semantic_scholarresearch

Multi-consensus task implementation of fractional-order multi-agent system with distributed adaptive parameter controller

λΆ„μ‚° μ œμ–΄ 기법을 μ΄μš©ν•œ 닀쀑 μ—μ΄μ „νŠΈ ν•©μ˜ 문제 ν•΄κ²°

semantic_scholarresearch

Fully actuated system approach based adaptive formation control for heterogeneous multi-agent systems with asymmetric full-state constraints under fixed and switching topologies

이질적 μ—μ΄μ „νŠΈμ˜ 동적 νŽΈλŒ€ μ œμ–΄λ₯Ό μœ„ν•œ 적응적 μ œμ–΄ 방법λ‘