-
Introducing OpenDevin CodeAct 1.0, a new State-of-the-art in Coding Agents
By Xingyao Wang, Bowen Li, and Graham Neubig.
-
How should LLM agents best interact with our world?
Blog for paper "Executable Code Actions Elicit Better LLM Agents"
-
A new benchmark tailored for LLMs' multi-turn interactions
Blog for ICLR 2024 paper "MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedback"