TY  - RPRT
TI  - LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraints
AU  - Thomas Palmeira Ferraz
AU  - Kartik Mehta
AU  - Yu-Hsiang Lin
AU  - Haw-Shiuan Chang
AU  - Shereen Oraby
AU  - Sijia Liu
AU  - Vivek Subramanian
AU  - Tagyoung Chung
AU  - Mohit Bansal
AU  - Nanyun Peng
PY  - 2026
DO  - 10.18653/v1/2024.findings-emnlp.458
UR  - https://arxiv.org/abs/2410.06458
ID  - 2410.06458
ER  -