Reflexion: Language Agents with Verbal Reinforcement Learning

Noah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan, Shunyu Yao

🏛 Institutions: Northeastern University, Massachusetts Institute of Technology, Princeton University
📅 Date: March 20, 2023
📑 Publisher: NeurIPS 2023
💻 Env
🔑 Keywords: framework learning verbal reinforcement learning reflexion

TLDR

Introduces Reflexion, a framework where language agents learn from feedback by writing verbal reflections into episodic memory instead of updating model weights. The method is broad rather than GUI-specific, but it matters here because many later web and computer-use agents inherit its reflection-and-memory pattern.

Open paper Edit on GitHub Report issue