Chain of Draft: Thinking Faster by Writing Less
This addresses the problem of high computational cost and latency in LLM reasoning for AI researchers and practitioners, offering a more efficient alternative to existing methods.
The paper tackles the inefficiency of verbose reasoning in LLMs by proposing Chain of Draft (CoD), a paradigm that generates minimal intermediate thoughts, achieving comparable or better accuracy than Chain-of-Thought while using only 7.6% of the tokens.
Large Language Models (LLMs) have demonstrated remarkable performance in solving complex reasoning tasks through mechanisms like Chain-of-Thought (CoT) prompting, which emphasizes verbose, step-by-step reasoning. However, humans typically employ a more efficient strategy: drafting concise intermediate thoughts that capture only essential information. In this work, we propose Chain of Draft (CoD), a novel paradigm inspired by human cognitive processes, where LLMs generate minimalistic yet informative intermediate reasoning outputs while solving tasks. By reducing verbosity and focusing on critical insights, CoD matches or surpasses CoT in accuracy while using as little as only 7.6% of the tokens, significantly reducing cost and latency across various reasoning tasks. Our code and data are available at https://github.com/sileix/chain-of-draft.