Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Mentor-KD: Making Small Language Models Better Multi-step Reasoners

About

Large Language Models (LLMs) have displayed remarkable performances across various complex tasks by leveraging Chain-of-Thought (CoT) prompting. Recently, studies have proposed a Knowledge Distillation (KD) approach, reasoning distillation, which transfers such reasoning ability of LLMs through fine-tuning language models of multi-step rationales generated by LLM teachers. However, they have inadequately considered two challenges regarding insufficient distillation sets from the LLM teacher model, in terms of 1) data quality and 2) soft label provision. In this paper, we propose Mentor-KD, which effectively distills the multi-step reasoning capability of LLMs to smaller LMs while addressing the aforementioned challenges. Specifically, we exploit a mentor, intermediate-sized task-specific fine-tuned model, to augment additional CoT annotations and provide soft labels for the student model during reasoning distillation. We conduct extensive experiments and confirm Mentor-KD's effectiveness across various models and complex reasoning tasks.

Hojae Lee, Junho Kim, SangKeun Lee• 2024

Related benchmarks

TaskDatasetResultRank
Commonsense ReasoningStrategyQA
Accuracy60
208
Commonsense ReasoningCSQA
CSQA Accuracy66.9
195
Symbolic ReasoningLast Letter Concatenation
Accuracy61
68
Logical reasoningShuffled Objects
Accuracy88.3
39
ReasoningBBH
BBH Score44.3
39
Symbolic ReasoningLast Letter
Accuracy0.638
31
Showing 6 of 6 rows

Other info

Follow for update