Problem-Solving Logic Guided Curriculum In-Context Learning for LLMs Complex Reasoning

TL;DR

This paper introduces a problem-solving logic guided curriculum in-context learning method that improves large language models' complex reasoning by selecting and ordering examples based on problem-solving logic and difficulty, outperforming previous approaches.

Contribution

The study proposes a novel curriculum ICL strategy guided by problem-solving logic, including a fine-tuned model and a new instruction set for better example selection and ordering.

Findings

Outperforms previous ICL methods in benchmarks

Enhances reasoning capabilities of LLMs

Improves efficiency of in-context learning

Abstract

In-context learning (ICL) can significantly enhance the complex reasoning capabilities of large language models (LLMs), with the key lying in the selection and ordering of demonstration examples. Previous methods typically relied on simple features to measure the relevance between examples. We argue that these features are not sufficient to reflect the intrinsic connections between examples. In this study, we propose a curriculum ICL strategy guided by problem-solving logic. We select demonstration examples by analyzing the problem-solving logic and order them based on curriculum learning. Specifically, we constructed a problem-solving logic instruction set based on the BREAK dataset and fine-tuned a language model to analyze the problem-solving logic of examples. Subsequently, we selected appropriate demonstration examples based on problem-solving logic and assessed their difficulty according to the number of problem-solving steps. In accordance with the principles of curriculum learning, we ordered the examples from easy to hard to serve as contextual prompts. Experimental results on multiple benchmarks indicate that our method outperforms previous ICL approaches in terms of performance and efficiency, effectively enhancing the complex reasoning capabilities of LLMs. Our project will be released at https://github.com/maxuetao/CurriculumICL