Beyond Output Critique: Self-Correction via Task Distillation

TL;DR

The paper introduces SELF-THOUGHT, a framework that enhances LLM self-correction by using task abstraction to guide solution refinement, improving accuracy and robustness across models.

Contribution

It proposes a novel task distillation approach that transfers structured task templates from large to small models, enabling more reliable self-correction without extensive fine-tuning.

Findings

Improves accuracy and robustness in diverse reasoning tasks.

Enables small models to benefit from large model templates.

Reduces error propagation through task abstraction.

Abstract

Large language models (LLMs) have shown promising self-correction abilities, where iterative refinement improves the quality of generated responses. However, most existing approaches operate at the level of output critique, patching surface errors while often failing to correct deeper reasoning flaws. We propose SELF-THOUGHT, a framework that introduces an intermediate step of task abstraction before solution refinement. Given an input and an initial response, the model first distills the task into a structured template that captures key variables, constraints, and problem structure. This abstraction then guides solution instantiation, grounding subsequent responses in a clearer understanding of the task and reducing error propagation. Crucially, we show that these abstractions can be transferred across models: templates generated by larger models can serve as structured guides for smaller LLMs, which typically struggle with intrinsic self-correction. By reusing distilled task structures, smaller models achieve more reliable refinements without heavy fine-tuning or reliance on external verifiers. Experiments across diverse reasoning tasks demonstrate that SELF-THOUGHT improves accuracy, robustness, and generalization for both large and small models, offering a scalable path toward more reliable self-correcting language systems.