LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraints
Thomas Palmeira Ferraz, Kartik Mehta, Yu-Hsiang Lin, Haw-Shiuan Chang, Shereen Oraby, Sijia Liu, Vivek Subramanian, Tagyoung Chung, Mohit Bansal, and Nanyun Peng, in Proceedings of the Findings of ACL at The 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP-Findings), 2024.
Abstract
We investigate LLMs’ capability in following multi-constrained instructions, introducing the Decompose, Critique, and Refine (DeCRIM) self-correction pipeline. This approach significantly enhances the ability of LLMs to handle complex constraints, and our experiments demonstrate substantial improvements in instruction adherence across multiple evaluation metrics.
Bib Entry
@inproceedings{ferraz2024llm,
title = {LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraints},
author = {Ferraz, Thomas Palmeira and Mehta, Kartik and Lin, Yu-Hsiang and Chang, Haw-Shiuan and Oraby, Shereen and Liu, Sijia and Subramanian, Vivek and Chung, Tagyoung and Bansal, Mohit and Peng, Nanyun},
booktitle = {Proceedings of the Findings of ACL at The 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP-Findings)},
year = {2024}
}