用人类思维步骤引导大模型解题，提升复杂任务表现。
Unlocking Structured Thinking in Language Models with Cognitive Prompting
- 模拟人类思考过程，分步完成目标澄清、分解等操作
- 在GSM8K上相比标准问答显著提效，小模型也明显受益
- 适合需要严谨推理的数学题、逻辑题场景
我们提出认知提示（cognitive prompting），一种通过结构化、类人认知操作（如目标澄清、分解、过滤、抽象和模式识别）引导大语言模型（LLMs）解决问题的新方法。通过系统性分步推理，该方法使模型更高效地处理复杂多步任务。我们设计了三种变体：确定性序列、自适应选择序列的变体，以及利用生成正确答案作为少样本链式思维提示的混合变体。在LLaMA、Gemma~2和Qwen三种模型的两个尺寸上，于算术推理基准GSM8K上的实验表明，认知提示显著优于标准问答方式。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。