Shunyu Yao
First author of ReAct, Tree of Thoughts and the τ-bench agent benchmark; language-agent researcher
Shunyu Yao studies language agents. Yao is the first-listed author of ReAct (2022), which interleaves reasoning traces and actions in language models and was described on the Google Research blog by Yao as a Student Researcher on the Brain Team, and of Tree of Thoughts, WebShop and τ-bench, a benchmark of conversations between a simulated user and a tool-using agent. Yao's homepage lists OpenAI's Computer-Using Agent (CUA) and Deep Research among Yao's selected work, and a PhD thesis titled "Language Agents: From Next-Token Prediction to Digital Automation".1324
Work
τ-benchOpen-source code
First author, with Noah Shinn, Pedram Razavi and Karthik Narasimhan, of "τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains" (2024), which emulates conversations between a language-model-simulated user and an agent with domain-specific API tools and policy guidelines. The sierra-research/tau-bench repository is MIT-licensed.451
ReAct (external site: arxiv.org)Published research
First author of "ReAct: Synergizing Reasoning and Acting in Language Models" (2022), which has language models generate reasoning traces and task-specific actions in an interleaved manner.32
Other documented roles
Roles the sources document but that are not confirmed as current.
- Researcher, OpenAI1
- Student Researcher, Google Research (Brain Team)2
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.