; they would find it difficult to create suitable prompts for directing a communicative agent to develop a trading application. This predicament is raising a crucial question: can we replace human intervention with an autonomous communicative agent capable of steering the conversation toward task completion without any human supervision?
문제제기 : trading application을 만들도록 지시하는 prompts를 생성하는 건 어려움. 왜냐하면 이해를 잘못하기 때문. 이 상태는 중요한 질문으로 올라옴 : 사람의 간섭을 자동 대화 agent로 대체하여 대화를 주도할 수 있을까?
This paper explores the potential of building scalable techniques to facilitate autonomous cooperation among communicative agents and provide insight into their “cognitive” processes. To address these issues, we propose a novel cooperative agent framework named role-playing to automate cooperation between communicative agents
이 논문은 인지하는 단계에 통찰력을 제공하고 대화 agent에 자동 협력 기능을 위한 scalable 기술을 세움. 이 문제는 저자가 communicative agents 사이에 자동 cooperation으로 role-playing이라는 framework를 통합함.
Our proposed framework is a novel role-playing approach for studying multiple communicative agents. Specifically, we concentrate on task-oriented role-playing that involves one AI assistant and one AI user. After the multi-agent system receives a preliminary idea and the role assignment from human users, a task-specifier agent will provide a detailed description to make the idea specific and then the AI assistant and AI user will cooperate on completing the specified task through multi-turn conversations until the AI user determines the task is done. The AI user is responsible for giving instructions to the AI assistant and directing the conversation toward task completion. On the other hand, the AI assistant is designed to follow the instructions from the AI user and respond with specific solutions.
role-playing framework. AI assistance와 AI user로 나뉨. 아이디어에 대한 설명을 제공하고 AI user가 업무를 끝낼때까지 여러 대화로 계속 진행해감. AI user는 AI assistant한테 방향과 업무 성과를 알려줌. 반대로 AI assistant는 특정 문제에 응답과 AI user로부터 지시를 따라 design함.
Human Input and Task Specifying
The role-playing session will be instantiated from an idea and selected roles by humans. As an example in Figure 1, a human has a preliminary idea to develop a trading bot for the stock market.
What is needed is only to designate the potential roles that can implement the idea.
After the idea and roles are determined, the task specifier agent will brainstorm a specific task that the AI Assistant role can help with the AI user role to complete based on the input idea
role-playing session은 사람들의 역할 선택과 아이디어로부터 추상적인 것을 구체적으로 설명하게 됨. 하지마 만약 ㅇ아이디어가 가르킬 필요가 있다면? idea와 roles이 결정된 후, task specifier agent는 입력한 아이디어 기반으로 AI assistant는 AI user 역할을 도울 수 있는 구체적인 업무를 brain storm함. \
t. Therefore, the task specifier agent performs as an enhanced imagination module for the idea implementation.
task specifer agent는 아이디어 구현을 위한 향상된 imigination module로 수행됨.
AI Assistant-User Role Assignment.
. When the system message is passed to those models respectively, we obtain A ← F PA 1 and U ← F PU 2 which are referred to as the assistant and user agents respectively.
system message는 large-scale auto-regressive language models $F_1$, $F_2$를 통과하하여 $A,U$ 값을 얻어 assistant와 agent 각각 언급됨. Fig1에 assistant 역할을 python programmer, user는 stock trader로 나온 것으로 이미 추측 할 수 있음.
Conversation Towards Task-Solving.