Unlocking Structured Thinking in Language Models with Cognitive Prompting

TL;DR

This paper introduces cognitive prompting, a method that guides large language models through structured, human-like reasoning steps to improve their ability to solve complex tasks more effectively.

Contribution

It presents a novel cognitive prompting approach with three variants, enhancing LLM reasoning by mimicking human cognitive operations for better problem-solving.

Findings

Cognitive prompting significantly improves LLM performance on GSM8K.

The hybrid variant with few-shot chain-of-thought achieves notable accuracy gains.

Experiments on multiple models demonstrate the effectiveness of the approach.

Abstract

We propose cognitive prompting as a novel approach to guide problem-solving in large language models (LLMs) through structured, human-like cognitive operations, such as goal clarification, decomposition, filtering, abstraction, and pattern recognition. By employing systematic, step-by-step reasoning, cognitive prompting enables LLMs to tackle complex, multi-step tasks more efficiently. We introduce three variants: a deterministic sequence of cognitive operations, a self-adaptive variant in which the LLM dynamically selects the sequence of cognitive operations, and a hybrid variant that uses generated correct solutions as few-shot chain-of-thought prompts. Experiments with LLaMA, Gemma~2, and Qwen models in each two sizes on the arithmetic reasoning benchmark GSM8K demonstrate that cognitive prompting significantly improves performance compared to standard question answering.