Meta-Reasoning: Semantics-Symbol Deconstruction for Large Language Models

TL;DR

This paper introduces Meta-Reasoning, a linguistic approach that enables large language models to deconstruct semantic information into symbolic representations, improving reasoning capabilities across diverse tasks without relying on formal programming languages.

Contribution

It proposes a novel meta-reasoning method that enhances LLM reasoning by focusing on semantic-symbolic deconstruction, broadening applicability beyond formal language conversion.

Findings

Significantly improves reasoning accuracy across multiple datasets

Enhances out-of-domain generalization and output stability

Outperforms Chain-of-Thought techniques in experiments

Abstract

Neural-symbolic methods have demonstrated efficiency in enhancing the reasoning abilities of large language models (LLMs). However, existing methods mainly rely on syntactically mapping natural languages to complete formal languages like Python and SQL. Those methods require that reasoning tasks be convertible into programs, which cater to the computer execution mindset and deviate from human reasoning habits. To broaden symbolic methods' applicability and adaptability in the real world, we propose the Meta-Reasoning from a linguistic perspective. This method empowers LLMs to deconstruct reasoning-independent semantic information into generic symbolic representations, thereby efficiently capturing more generalized reasoning knowledge. We conduct extensive experiments on more than ten datasets encompassing conventional reasoning tasks like arithmetic, symbolic, and logical reasoning, and the more complex interactive reasoning tasks like theory-of-mind reasoning. Experimental results demonstrate that Meta-Reasoning significantly enhances in-context reasoning accuracy, learning efficiency, out-of-domain generalization, and output stability compared to the Chain-of-Thought technique. Code and data are publicly available at \url{https://github.com/Alsace08/Meta-Reasoning}.