Prompting Science Report 2: The Decreasing Value of Chain of Thought in Prompting

TL;DR

This report critically examines Chain-of-Thought prompting in large language models, revealing its variable effectiveness across different models and tasks, often increasing computational cost without consistent accuracy improvements.

Contribution

It provides a nuanced analysis of CoT prompting, showing its limited benefits for models with inherent reasoning abilities and highlighting increased costs and variability.

Findings

CoT improves performance mainly in non-reasoning models

Recent models often perform implicit CoT reasoning without prompts

CoT increases token usage and response time without guaranteed accuracy gains

Abstract

This is the second in a series of short reports that seek to help business, education, and policy leaders understand the technical details of working with AI through rigorous testing. In this report, we investigate Chain-of-Thought (CoT) prompting, a technique that encourages a large language model (LLM) to "think step by step" (Wei et al., 2022). CoT is a widely adopted method for improving reasoning tasks, however, our findings reveal a more nuanced picture of its effectiveness. We demonstrate two things: - The effectiveness of Chain-of-Thought prompting can vary greatly depending on the type of task and model. For non-reasoning models, CoT generally improves average performance by a small amount, particularly if the model does not inherently engage in step-by-step processing by default. However, CoT can introduce more variability in answers, sometimes triggering occasional errors in questions the model would otherwise get right. We also found that many recent models perform some form of CoT reasoning even if not asked; for these models, a request to perform CoT had little impact. Performing CoT generally requires far more tokens (increasing cost and time) than direct answers. - For models designed with explicit reasoning capabilities, CoT prompting often results in only marginal, if any, gains in answer accuracy. However, it significantly increases the time and tokens needed to generate a response.