Simple Agent
Simple agent research focuses on designing and optimizing autonomous systems capable of completing tasks, often by integrating large language models (LLMs) with tools and other agents. Current efforts concentrate on developing unified frameworks for agent architecture, automating the design process through meta-agent learning algorithms, and ensuring agent safety and controllability, including methods for shutdown. This field is significant for advancing artificial intelligence, particularly in areas like automated root cause analysis, assistive technologies for individuals with cognitive disabilities, and efficient resource management in complex systems such as satellite networks.
Papers
Game-theoretic LLM: Agent Workflow for Negotiation Games
Wenyue Hua, Ollie Liu, Lingyao Li, Alfonso Amayuelas, Julie Chen, Lucas Jiang, Mingyu Jin, Lizhou Fan, Fei Sun, William Wang, Xintong Wang, Yongfeng Zhang
A Taxonomy of AgentOps for Enabling Observability of Foundation Model based Agents
Liming Dong, Qinghua Lu, Liming Zhu