https://arxiv.org/pdf/2308.08155
가장 큰 특징 : customizable, conversable, conversation programming
1.introduction
세가지 이유
1) 지금의 LLM은 the ability to incorporate feedback을 가짐
2) single LLM can exhibit a broad range of capabilities (특히 정확한 프롬프트와 inference환경으로 configured일때), conversations between differently configured agents can help combine these broad LLM capabilities in a modular and complementary manner
3) Third, LLMs have demonstrated ability to solve complex tasks when the tasks are broken into simpler subtasks
복잡한 과제를 decompose시 좋은 성능을 보임
이를 위해 두가지 질문에 답할 필요성
(1) How can we design individual agents that are capable, reusable, customizable, and effective in multi-agent collaboration? 멀티 에이젼트 환경에서 각 agent를 어떻게 디자인해야하는가
(2) How can we develop a straightforward, unified interface that can accommodate a wide range of agent conversation patterns?
에이젼트간의 대화에 적용할 수있는 interface를 어떻게 개발해야하는가
agent의 role에는 코드를 작성, 코드 실행, 인간의 피드백을 시스템에 통합(wire in), output을 validate 등이 있다
2.1 conversable agent
2.2 conversation programming
여섯개의 application
'agent > multi - agent' 카테고리의 다른 글
mind search 논문리뷰 (0) | 2024.08.01 |
---|---|
CRITIC: LARGE LANGUAGE MODELS CAN SELFCORRECT WITH TOOL-INTERACTIVE CRITIQUING 논문리뷰 (0) | 2024.07.29 |
metagpt 논문리뷰 (0) | 2024.07.28 |
LEGO: A Multi-agent Collaborative Framework with Role-playing andIterative Feedback for Causality Explanation Generation 논문리뷰 (0) | 2024.07.26 |
Debating with More Persuasive LLMs Leads to More Truthful Answers 논문리뷰 (0) | 2024.07.24 |