Look Before You Leap: Towards Decision-Aware and Generalizable Tool-Usage for Large Language Models

TL;DR

This paper introduces DEER, a decision-aware and generalizable framework for improving open-source LLMs' tool-usage capabilities, enabling better handling of diverse and unseen tools and queries.

Contribution

The paper proposes a novel framework that enhances LLMs' decision-making and generalizability in tool-usage through automatic sample generation and strategic tool sampling.

Findings

DEER significantly outperforms baselines on various datasets.

The framework improves LLMs' flexibility in handling diverse queries.

DEER enhances generalization to unseen tools.

Abstract

Tool-augmented large language models (LLMs) are attracting widespread attention when accessing up-to-date knowledge and alleviating hallucination issues. Nowadays, advanced closed-source LLMs (e.g., ChatGPT) have demonstrated surprising tool-usage capabilities through prompting and in-context learning techniques. To empower the capabilities of open-source LLMs (e.g., LLaMA) in manipulating tools, current efforts focus on either template-driven or token-triggered tool-usage. However, the former hampers LLMs' flexibility to address diverse user's queries due to constrained tool interactions, while the latter limits the generalizability when engaging with new tools, since tool-usage learning is based on task- and tool-specific datasets. To alleviate these concerns, in this paper, we propose a decision-aware and generalizable tool-usage framework (DEER). Specifically, we first construct the tool-usage samples with multiple decision branches via an automatic generation pipeline, thereby inspiring the decision-making awareness of LLMs under diverse scenarios. Meanwhile, we propose a novel tool sampling strategy to enhance the generalizability of LLMs over unseen tools. Extensive experiments demonstrate that our proposed DEER is effective and significantly outperforms baselines across various datasets.