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

Reinforcement-aware Knowledge Distillation for LLM Reasoning

About

Reinforcement learning (RL) post-training has recently driven major gains in long chain-of-thought reasoning large language models (LLMs), but the high inference cost of such models motivates distillation into smaller students. Most existing knowledge distillation (KD) methods are designed for supervised fine-tuning (SFT), relying on fixed teacher traces or teacher-student Kullback-Leibler (KL) divergence-based regularization. When combined with RL, these approaches often suffer from distribution mismatch and objective interference: teacher supervision may not align with the student's evolving rollout distribution, and the KL regularizer can compete with reward maximization and require careful loss balancing. To address these issues, we propose RL-aware distillation (RLAD), which performs selective imitation during RL -- guiding the student toward the teacher only when it improves the current policy update. Our core component, Trust Region Ratio Distillation (TRRD), replaces the teacher-student KL regularizer with a PPO/GRPO-style likelihood-ratio objective anchored to a teacher--old-policy mixture, yielding advantage-aware, trust-region-bounded distillation on student rollouts and naturally balancing exploration, exploitation, and imitation. Across diverse logic reasoning and math benchmarks, RLAD consistently outperforms offline distillation, standard GRPO, and KL-based on-policy teacher-student knowledge distillation.

Zhaoyang Zhang, Shuli Jiang, Yantao Shen, Yuting Zhang, Dhananjay Ram, Shuo Yang, Zhuowen Tu, Wei Xia, Stefano Soatto• 2026

Related benchmarks

TaskDatasetResultRank
Mathematical ReasoningMATH500 (test)--
514
Mathematical ReasoningAIME 2024 (test)--
159
Mathematical ReasoningAIME 2025 (test)
Pass@1 Rate25.1
63
Mathematical ReasoningAIME 2024
Pass@147.8
39
Mathematical ReasoningAIME 2025
Pass@131.7
39
Mathematical ReasoningMathematical Reasoning Aggregate
Average Score59.1
26
Mathematical ReasoningAMC 2024
Pass@149.8
19
Math ReasoningMATH500
Pass@168.6
18
Logical reasoningK&K Logistics (test)
PPL20.98
18
Math ReasoningAMC 2023 (test)
Pass@178.2
16
Showing 10 of 12 rows

Other info

Follow for update