Teaching Small Language Models to Reason for Knowledge-Intensive Multi-Hop Question Answering

Xiang Li, Shizhu He, Fangyu Lei, JunYang JunYang, Tianhuang Su, Kang Liu, Jun Zhao


Abstract
Large Language Models (LLMs) can teach small language models (SLMs) to solve complex reasoning tasks (e.g., mathematical question answering) by Chain-of-thought Distillation (CoTD). Specifically, CoTD fine-tunes SLMs by utilizing rationales generated from LLMs such as ChatGPT. However, CoTD has certain limitations that make it unsuitable for knowledge-intensive multi-hop question answering: 1) SLMs have a very limited capacity in memorizing required knowledge compared to LLMs. 2) SLMs do not possess the same powerful integrated abilities in question understanding and knowledge reasoning as LLMs. To address the above limitations, we introduce Decompose-and-Response Distillation (D&R Distillation), which distills two student models, namely Decomposer and Responser separately. The two models solve a knowledge-intensive multi-hop question through an interactive process of asking and answering subquestions. Our method offers two advantages: 1) SLMs have the capability to access external knowledge to address subquestions, which provides more comprehensive knowledge for multi-hop questions. 2) By employing simpler subquestions instead of complex CoT reasoning, SLMs effectively mitigate task complexity and decrease data prerequisites. Experimental results on three knowledge-intensive multi-hop question answering datasets demonstrate that D&R Distillation can surpass previous CoTD methods, even with much less training data.
Anthology ID:
2024.findings-acl.464
Volume:
Findings of the Association for Computational Linguistics: ACL 2024
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
7804–7816
Language:
URL:
https://aclanthology.org/2024.findings-acl.464
DOI:
10.18653/v1/2024.findings-acl.464
Bibkey:
Cite (ACL):
Xiang Li, Shizhu He, Fangyu Lei, JunYang JunYang, Tianhuang Su, Kang Liu, and Jun Zhao. 2024. Teaching Small Language Models to Reason for Knowledge-Intensive Multi-Hop Question Answering. In Findings of the Association for Computational Linguistics: ACL 2024, pages 7804–7816, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Teaching Small Language Models to Reason for Knowledge-Intensive Multi-Hop Question Answering (Li et al., Findings 2024)
Copy Citation:
PDF:
https://preview.aclanthology.org/dois-2013-emnlp/2024.findings-acl.464.pdf