# Can Small Language Models Help Large Language Models Reason Better?: LM-Guided Chain-of-Thought

Source: [Can Small Language Models Help Large Language Models Reason Better?: LM-Guided Chain-of-Thought](<https://www.sota2.com/research/paper/68fa72d21e209131cea58b7a>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T02:34:49.549Z

Authors: Jooyoung Lee, Fan Yang, Thanh Tran, Qian Hu, Emre Barut, Kai-Wei Chang, Chengwei Su

Published: 2024-04-04T12:46:37.000Z

## Abstract

We introduce a novel framework, LM-Guided CoT, that leverages a lightweight (i.e., \<1B) language model (LM) for guiding a black-box large (i.e., \>10B) LM in reasoning tasks. Specifically, the lightweight LM first generates a rationale for each input instance. The Frozen large LM is then prompted to predict a task output based on the rationale generated by the lightweight LM. Our approach is resource-efficient in the sense that it only requires training the lightweight LM. We optimize the model through 1) knowledge distillation and 2) reinforcement learning from rationale-oriented and task-oriented reward signals. We assess our method with multi-hop extractive question answering (QA) benchmarks, HotpotQA, and 2WikiMultiHopQA. Experimental results show that our approach outperforms all baselines regarding answer prediction accuracy. We also find that reinforcement learning helps the model to produce higher-quality rationales with improved QA performance.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2404.03414>)

[PDF](<https://arxiv.org/pdf/2404.03414.pdf>)
