Papers
arxiv:2505.13975

DRP: Distilled Reasoning Pruning with Skill-aware Step Decomposition for Efficient Large Reasoning Models

Published on May 20
Authors:
,

Abstract

Distilled Reasoning Pruning enhances efficiency and accuracy in mathematical reasoning tasks by reducing verbose reasoning traces through a combination of pruning and distillation.

AI-generated summary

While Large Reasoning Models (LRMs) have demonstrated success in complex reasoning tasks through long chain-of-thought (CoT) reasoning, their inference often involves excessively verbose reasoning traces, resulting in substantial inefficiency. To address this, we propose Distilled Reasoning Pruning (DRP), a hybrid framework that combines inference-time pruning with tuning-based distillation, two widely used strategies for efficient reasoning. DRP uses a teacher model to perform skill-aware step decomposition and content pruning, and then distills the pruned reasoning paths into a student model, enabling it to reason both efficiently and accurately. Across several challenging mathematical reasoning datasets, we find that models trained with DRP achieve substantial improvements in token efficiency without sacrificing accuracy. Specifically, DRP reduces average token usage on GSM8K from 917 to 328 while improving accuracy from 91.7% to 94.1%, and achieves a 43% token reduction on AIME with no performance drop. Further analysis shows that aligning the reasoning structure of training CoTs with the student's reasoning capacity is critical for effective knowledge transfer and performance gains.

Community

Sign up or log in to comment

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2505.13975 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2505.13975 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2505.13975 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.