cs.CLNov 14, 2026

B-OPSD: On-Policy Self-Distillation with Self-Generated Problem Briefs

Authors: Zhi Xu, Yinqi Wu, Feng Zhang, Xiaotong Zhang, Hong Yu, Han Liu

Organizations: Dalian University of Technology, Dalian, China

Abstract

On-Policy Self-Distillation (OPSD) has recently attracted interest because it provides dense teacher guidance without a separate, often larger teacher model. However, OPSD conditions its teacher on ground-truth solutions, which limits its use when such solutions are unavailable. We propose On-Policy Self-Distillation with Self-Generated Problem Briefs (B-OPSD). Using a prompt template, the initial model or a saved student checkpoint generates a problem brief from the question alone. The prompt asks for relevant principles, applicability conditions, and likely pitfalls, but not a final answer or a complete solution. A frozen teacher uses the brief as privileged context to guide the current student's on-policy rollouts through token-level distribution matching. B-OPSD uses no ground-truth solutions during training. Results on AIME 2024, AIME 2025, and HMMT 2025 show that both B-OPSD variants achieve higher average accuracy than OPSD at each of three Qwen3 model sizes. Code is available at https://github.com/ZevineXu/B-OPSD

Figures & tables

Explore similar work

CardsList
  1. DualOPSD: Closing the Feedback Loop in On-Policy Self-Distillation

    Aug 26, 2026Yutong Chen, Guangfu Guo, Zhichao Xu +1Adaptive Knowledge DistillationOn-Policy Self-Distillation

  2. Privileged Solutions or Context-Induced Teacher Behavior? Dissecting On-Policy Self-Distillation

    Aug 10, 2026Yuki Ichihara, Naoto Iwase, Mohammad Atif Quamar +1On-Policy Self-DistillationLanguage Model Distillation

  3. Learning from a Thoughtful Teacher: Adaptive On-Policy Self-Distillation for Mathematical Reasoning

    Sep 26, 2026Jiacheng Du, Weiwei Xie, Tianyi Du +3On-Policy Self-DistillationCoT Distillation