First Heuristic Then Rational: Dynamic Use of Heuristics in Language Model Reasoning

TL;DR

This paper investigates how language models dynamically switch between heuristic and rational strategies during multi-step reasoning, revealing a shift from heuristic reliance early on to more rational processing near the conclusion.

Contribution

It uncovers the dynamic use of heuristics versus rational strategies in language model reasoning, highlighting the limited backtracking ability and adaptive strategy combination.

Findings

LMs rely more on heuristics early in reasoning

Reliance on heuristics decreases as reasoning progresses

Limited backtracking capability in multi-step reasoning

Abstract

Multi-step reasoning instruction, such as chain-of-thought prompting, is widely adopted to explore better language models (LMs) performance. We report on the systematic strategy that LMs employ in such a multi-step reasoning process. Our controlled experiments reveal that LMs rely more heavily on heuristics, such as lexical overlap, in the earlier stages of reasoning, where more reasoning steps remain to reach a goal. Conversely, their reliance on heuristics decreases as LMs progress closer to the final answer through multiple reasoning steps. This suggests that LMs can backtrack only a limited number of future steps and dynamically combine heuristic strategies with rationale ones in tasks involving multi-step reasoning.