Call Me When Necessary: LLMs can Efficiently and Faithfully Reason over Structured Environments
This work addresses efficient and faithful reasoning for question answering over structured data, representing an incremental improvement over previous LLM-based methods.
The paper tackles the problem of multi-hop reasoning over structured environments like knowledge graphs and tables by proposing Reasoning-Path-Editing (Readi), a framework where LLMs generate and edit reasoning paths only when necessary, resulting in significant performance gains such as 9.1% Hit@1 improvement on WebQSP and 14.9% boost on CWQ compared to vanilla LLMs.
Large Language Models (LLMs) have shown potential in reasoning over structured environments, e.g., knowledge graph and table. Such tasks typically require multi-hop reasoning, i.e., match natural language utterance with instances in the environment. Previous methods leverage LLMs to incrementally build a reasoning path, where the LLMs either invoke tools or pick up schemas by step-by-step interacting with the environment. We propose Reasoning-Path-Editing (Readi), a novel framework where LLMs can efficiently and faithfully reason over structured environments. In Readi, LLMs initially generate a reasoning path given a query, and edit the path only when necessary. We instantiate the path on structured environments and provide feedback to edit the path if anything goes wrong. Experimental results on three KGQA and two TableQA datasets show the effectiveness of Readi, significantly surpassing previous LLM-based methods (by 9.1% Hit@1 on WebQSP, 12.4% on MQA-3H and 9.5% on WTQ), comparable with state-of-the-art fine-tuned methods (67% on CWQ and 74.7% on WebQSP) and substantially boosting the vanilla LLMs (by 14.9% on CWQ). Our code will be available on https://aka.ms/readi.