Making Language Models Better Tool Learners with Execution Feedback
Preprint
Abstract
Tools serve as pivotal interfaces that enable humans to understand and reshape the environment. With the advent of foundation models, AI systems can utilize tools to expand their capabilities and interact with the real world. Existing tool learning methodologies, encompassing supervised fine-tuning and prompt engineering approaches, often induce large language models to utilize tools indiscriminately, as complex tasks often exceed their own competencies. However, introducing tools for simple tasks, which the models themselves can readily resolve, can inadvertently propagate errors rather than enhance performance. This leads to the research question: can we teach language models when and how to use tools? To meet this need, we propose Tool leaRning wIth exeCution fEedback (TRICE), a two-stage end-to-end framework that enables the model to continually learn through feedback derived from tool execution, thereby learning when and how to use tools effectively. Experimental results, backed by further analysis, show that TRICE can make the large language model selectively use tools by improving the accuracy of tool usage while enhancing insufficient tool learning and mitigating excessive reliance on tools. Code is available at https://github.com/zjunlp/TRICE.Keywords
All Related Versions
This publication has 17 references indexed in Scilit:
- Open-vocabulary Queryable Scene Representations for Real World Planning2023
- ChatCoT: Tool-Augmented Chain-of-Thought Reasoning on Chat-based Large Language Models2023
- Verify-and-Edit: A Knowledge-Enhanced Chain-of-Thought Framework2023
- GLM: General Language Model Pretraining with Autoregressive Blank Infilling2022
- Are NLP Models really able to Solve Simple Math Word Problems?2021
- Mastering Atari, Go, chess and shogi by planning with a learned modelNature, 2020
- A Diverse Corpus for Evaluating and Developing English Math Word Problem Solvers2020
- MLQA: Evaluating Cross-lingual Extractive Question Answering2020
- Natural Questions: A Benchmark for Question Answering ResearchTransactions of the Association for Computational Linguistics, 2019
- Exploring the Limitations of Behavior Cloning for Autonomous Driving2019