From Dissonance to Orchestration: Teacher Intervention in On-Policy Distillation

arXiv:2609.37510 · cs.CL, cs.LG · Submitted 2026-09-29 · Read on arXiv

cs.CL, cs.LG

Submitted: 2026-09-29

Updated: 2026-09-29

Code: https://github.com/yhao-wang/MAESTRO

Terminology

• Accuracy — 29× in this paper · explained in 1 episode
• Benchmark — 25× in this paper · explained in 2 episodes
• Distillation — 21× in this paper · explained in 2 episodes
• On-Policy Distillation — 12× in this paper · explained in 3 episodes
• Aggregation — 11× in this paper · explained in 1 episode
• Ablation — 6× in this paper · explained in 1 episode
• Adaptation — 5× in this paper · explained in 1 episode
• Reinforcement Learning — 5× in this paper · explained in 4 episodes
• Alignment — 4× in this paper · explained in 1 episode
• Policy Disagreement Score — 4× in this paper
• GRPO — 3× in this paper · explained in 6 episodes
• Ablation study — 2× in this paper · explained in 1 episode
• Knowledge Distillation — 2× in this paper · explained in 11 episodes
• Positional Bias — 2× in this paper · explained in 2 episodes
• Adam — 1× in this paper · explained in 2 episodes
• Core — 1× in this paper · explained in 2 episodes
• Perplexity — 1× in this paper · explained in 3 episodes
• Regret — 1× in this paper · explained in 3 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Teacher-Student Gap — 1× in this paper · explained in 2 episodes

Sources

• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• Beyond Benchmarks: MathArena as an Evaluation Platform for Mathematics with LLMs
• Distilling the Knowledge in a Neural Network
• On the Direction of RLVR Updates for LLM Reasoning: Identification and Exploitation
• Trajectory-Refined Distillation
• Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Qwen3 Technical Report
• Pass the Baton: Trajectory-Relayed On-Policy Distillation
• Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement
• Less is More: Early Stopping Rollout for On-Policy Distillation

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering