Abstract
Language Models (LMs) have shown impressive performance in various natural language tasks. However, when it comes to natural language reasoning, LMs still face challenges such as hallucination, generating incorrect intermediate reasoning steps, and making mathematical errors. Recent research has focused on enhancing LMs through *self-improvement* using feedback. Nevertheless, existing approaches relying on a single generic feedback source fail to address the diverse error types found in LM-generated reasoning chains. In this work, we propose **Multi-Aspect Feedback**, an iterative refinement framework that integrates multiple feedback modules, including frozen LMs and external tools, each focusing on a specific error category. Our experimental results demonstrate the efficacy of our approach to addressing several errors in the LM-generated reasoning chain and thus improving the overall performance of an LM in several reasoning tasks. We see an improvement of up to 20% in Mathematical Reasoning and up to 18% in Logical Entailment.- Anthology ID:
- 2023.emnlp-main.407
- Volume:
- Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
- Month:
- December
- Year:
- 2023
- Address:
- Singapore
- Editors:
- Houda Bouamor, Juan Pino, Kalika Bali
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 6591–6616
- Language:
- URL:
- https://aclanthology.org/2023.emnlp-main.407
- DOI:
- 10.18653/v1/2023.emnlp-main.407
- Cite (ACL):
- Deepak Nathani, David Wang, Liangming Pan, and William Wang. 2023. MAF: Multi-Aspect Feedback for Improving Reasoning in Large Language Models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 6591–6616, Singapore. Association for Computational Linguistics.
- Cite (Informal):
- MAF: Multi-Aspect Feedback for Improving Reasoning in Large Language Models (Nathani et al., EMNLP 2023)
- PDF:
- https://preview.aclanthology.org/corrections-2024-05/2023.emnlp-main.407.pdf