小模型推理弱？用大模型精准补脑，效果翻倍
Guiding Reasoning in Small Language Models with LLM Assistance
- 小模型卡壳时，大模型只在关键步骤注入纠错提示
- 数学推理任务上，性能显著提升，接近大模型水平
- 适合想用小模型做复杂推理但算力有限的开发者
小语言模型（SLMs）在需要深度、多步逻辑推理的任务中表现受限。本文提出一种名为SMART的框架，通过大语言模型（LLM）对小模型进行选择性引导。受认知支架理论启发，SMART采用评分机制识别不确定的推理步骤，并仅在必要时注入由大模型生成的修正推理。将结构化推理建模为最优策略搜索，引导推理路径向正确解收敛，无需穷举采样。在数学推理数据集上的实验表明，有针对性的外部支持可显著提升性能，为小模型与大模型协同解决当前仅靠小模型无法完成的复杂推理任务开辟了新路径。
原文摘要 · Abstract (English)
The limited reasoning capabilities of small language models (SLMs) cast doubt on their suitability for tasks demanding deep, multi-step logical deduction. This paper introduces a framework called Small Reasons, Large Hints (SMART), which selectively augments SLM reasoning with targeted guidance from large language models (LLMs). Inspired by the concept of cognitive scaffolding, SMART employs a score-based evaluation to identify uncertain reasoning steps and injects corrective LLM-generated reasoning only when necessary. By framing structured reasoning as an optimal policy search, our approach steers the reasoning trajectory toward correct solutions without exhaustive sampling. Our experiments on mathematical reasoning datasets demonstrate that targeted external scaffolding significantly improves performance, paving the way for collaborative use of both SLM and LLM to tackle complex reasoning tasks that are currently unsolvable by SLMs alone.
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。