领域归一化多教师在线策略蒸馏方法
原标题：Paper page - Beyond Teacher Assignment: Domain-Normalized Multi-Teacher On-Policy Distillation
AI 摘要
该论文研究多教师在线策略蒸馏（MOPD），即让数学、编程、指令遵循等不同领域的专家模型共同指导学生模型。作者在三种规模的 Qwen3.5 模型上发现，MOPD 学生并未超过最佳单专家教学，且数学能力提升有限，原因是各领域反馈尺度不均衡，指令遵循反馈的离散度远大于数学反馈并主导了学生更新。为此提出领域归一化 MOPD（DN-MOPD），按各领域反馈的实测离散度重新缩放，在六个公开基准、三种规模、三个随机种子和两种答案长度限制下均优于 MOPD，并恢复了大部分数学增益。
正文节选
Beyond Teacher Assignment: Domain-Normalized Multi-Teacher On-Policy Distillation Abstract Reinforcement learning can turn one language model into several specialists, each excellent at a single skill such as mathematics, coding or following instructions, but users need one model with all of these skills. Multi-teacher on-policy distillation (MOPD) merges them by letting the specialists teach one student: the student answers each prompt, and the specialist for that prompt's domain gives feedback