工具检索桥：通过桥模型对模糊指令与检索器偏好进行对齐
Tool Retrieval Bridge: Aligning Vague Instructions with Retriever Preferences via Bridge Model
AI总结：
本文提出TRB方法，通过桥模型将模糊指令转化为更具体的指令，提升模糊指令下的工具检索性能，实验表明TRB在多个检索设置下显著提升检索效果。
摘要：
工具学习已成为大型语言模型（LLMs）解决现实挑战的有前景范式。由于工具数量庞大且更新频繁，选择所需的工具子集至关重要。然而，当前工具检索方法通常基于包含过于详细指令的学术基准（例如特定API名称和参数），而现实中的指令更模糊。这种差异会阻碍现实应用中的工具检索。本文首先构建了一个新的基准VGToolBench，以模拟人类模糊指令。基于此，我们进行了一系列初步分析，发现模糊指令确实损害了工具检索性能。为此，我们提出了一种简单而有效的工具检索桥（TRB）方法，以提升模糊指令下的工具检索性能。TRB的原理是引入一个桥模型，将模糊指令转换为更具体的指令，并缓解模糊指令与检索器偏好的差距。我们进行了大量实验，结果表明TRB有效缓解了模糊指令的歧义性，并在所有基线检索器上实现了一致且显著的改进。例如，在TRB的帮助下，BM25的相对提升可达111.51%，即平均NDCG分数从9.73增加到19.59。源代码和模型可在https://github.com/kfchenhn/TRB公开获取。
英文摘要：
Tool learning has emerged as a promising paradigm for large language models (LLMs) to address real-world challenges. Due to the extensive and irregularly updated number of tools, tool retrieval for selecting the desired tool subset is essential. However, current tool retrieval methods are usually based on academic benchmarks containing overly detailed instructions (e.g., specific API names and parameters), while real-world instructions are more vague. Such a discrepancy would hinder the tool retrieval in real-world applications. In this paper, we first construct a new benchmark, VGToolBench, to simulate human vague instructions. Based on this, we conduct a series of preliminary analyses and find that vague instructions indeed damage the performance of tool retrieval. To this end, we propose a simple-yet-effective Tool Retrieval Bridge (TRB) approach to boost the performance of tool retrieval for vague instructions. The principle of TRB is to introduce a bridge model to rewrite the vague instructions into more specific ones and alleviate the gap between vague instructions and retriever preferences.We conduct extensive experiments under multiple commonly used retrieval settings, and the results show that TRB effectively mitigates the ambiguity of vague instructions while delivering consistent and substantial improvements across all baseline retrievers. For example, with the help of TRB, BM25 achieves a relative improvement of up to 111.51%, i.e., increasing the average NDCG score from 9.73 to 19.59. The source code and models are publicly available at https://github.com/kfchenhn/TRB.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
