Agents & Tool Use¶
- Toolformer: Language Models Can Teach Themselves to Use Tools by Schick, Dwivedi-Yu, Dessì, Raber, Lomeli, Zettlemoyer, Cancedda, Scialom (2023)
[paper]— LMs learning when and how to call external tools. Key step toward agentic LMs. PDF - ReAct: Synergizing Reasoning and Acting in Language Models by Yao, Zhao, Yu, Du, Shafran, Narasimhan, Cao (2022)
[paper]— Interleaving reasoning traces and actions. The prompting pattern behind most LLM agents. PDF