ProgCo: Program Helps Self-Correction of Large Language Models
This addresses the challenge of unreliable self-correction in LLMs for complex reasoning, though it appears incremental by building on existing self-correction methods.
The paper tackles the problem of large language models failing to self-correct effectively in complex reasoning tasks by proposing ProgCo, which uses self-generated verification pseudo-programs and dual refinement, achieving effective self-correction as shown in experiments on three benchmarks.
Self-Correction aims to enable large language models (LLMs) to self-verify and self-refine their initial responses without external feedback. However, LLMs often fail to effectively self-verify and generate correct feedback, further misleading refinement and leading to the failure of self-correction, especially in complex reasoning tasks. In this paper, we propose Program-driven Self-Correction (ProgCo). First, program-driven verification (ProgVe) achieves complex verification logic and extensive validation through self-generated, self-executing verification pseudo-programs. Then, program-driven refinement (ProgRe) receives feedback from ProgVe, conducts dual reflection and refinement on both responses and verification programs to mitigate misleading of incorrect feedback in complex reasoning tasks. Experiments on three instruction-following and mathematical benchmarks indicate that ProgCo achieves effective self-correction, and can be further enhance performance when combined with real program tools. We release our code at https://github.com/songxiaoshuai/progco.