Chain of Logic: Rule-Based Reasoning with Large Language Models

TL;DR

This paper introduces Chain of Logic, a prompting method for large language models that improves rule-based reasoning, especially for complex compositional rules, by decomposing and recomposing logical elements inspired by legal reasoning frameworks.

Contribution

The paper presents a novel prompting technique, Chain of Logic, that enhances reasoning about compositional rules in language models by decomposing and recomposing logical elements.

Findings

Outperforms chain of thought and self-ask prompting methods

Effective across multiple rule-based reasoning tasks

Works with both open-source and commercial models

Abstract

Rule-based reasoning, a fundamental type of legal reasoning, enables us to draw conclusions by accurately applying a rule to a set of facts. We explore causal language models as rule-based reasoners, specifically with respect to compositional rules - rules consisting of multiple elements which form a complex logical expression. Reasoning about compositional rules is challenging because it requires multiple reasoning steps, and attending to the logical relationships between elements. We introduce a new prompting method, Chain of Logic, which elicits rule-based reasoning through decomposition (solving elements as independent threads of logic), and recomposition (recombining these sub-answers to resolve the underlying logical expression). This method was inspired by the IRAC (Issue, Rule, Application, Conclusion) framework, a sequential reasoning approach used by lawyers. We evaluate chain of logic across eight rule-based reasoning tasks involving three distinct compositional rules from the LegalBench benchmark and demonstrate it consistently outperforms other prompting methods, including chain of thought and self-ask, using open-source and commercial language models.