研究人员正在探索策略内蒸馏（OPD）的高级技术，通过结合多个“教师”模型来增强语言模型的能力。LEGO-OPD 和 SAKI 等新方法侧重于分解和路由教师信号，以在不降低性能的情况下改善视觉基础和推理。SCOUT 和 MAESTRO 等其他方法解决了教师-学生前缀不匹配的挑战，并优化了教师干预策略，以防止性能下降并确保跨不同任务的平衡能力集成。 AI
影响 策略内蒸馏的这些进展可能带来更强大、更专业的人工智能模型，从而提高复杂推理和多模态任务的性能。
排序理由 多篇研究论文介绍了策略内蒸馏的新技术。
AI 生成摘要 · Google Gemini · 来自 25 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2610.10460v1 Announce Type: new Abstract: Multi-teacher on-policy distillation (MOPD) is used in two settings. In common-domain composition, several teachers score each student rollout from one prompt domain and their signals form a single target; in routed-domain distillat…
arXiv:2610.10447v1 Announce Type: new Abstract: Reinforcement Learning (RL) from outcome rewards suffers from sparse supervision, particularly on difficult, long-horizon tasks where successful trajectories are rare and costly to generate. On-Policy Distillation (OPD) offers an at…
arXiv:2610.09639v1 Announce Type: new Abstract: On-policy distillation (OPD) strengthens language-model reasoning, yet whether students acquire new factual knowledge or compositional skill for multi-step reasoning remains unknown. We separate these capabilities using a controlled…
arXiv:2610.02324v1 Announce Type: cross Abstract: Foundation multimodal large language models are designed to support a broad spectrum of capabilities across diverse domains. Multi-teacher on-policy distillation (MOPD) provides an effective framework for consolidating domain-spec…
arXiv:2610.02678v1 Announce Type: new Abstract: On-policy distillation (OPD) combines student-generated rollouts with dense token-level supervision from a teacher, but providing such supervision for every rollout requires substantial teacher computation. We introduce Success-Refe…
arXiv:2610.02381v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student on the responses it generates. Existing LLM multi-teacher OPD transfers what specialists predict through their output distributions. We introduce Latent-MOPD, to our knowledge the first …
arXiv:2610.02179v1 Announce Type: new Abstract: Multi-teacher on-policy distillation (MOPD) aims to combine the strengths of RL-trained teachers in a single student, but how teacher signals affect parameter changes remains underexplored. We study Qwen3-1.7B with four domain teach…
On-policy distillation (OPD) has emerged as a widely used paradigm for post-training large language models, reducing the train--test mismatch of conventional distillation by supervising the student on its own generated trajectories. However, existing OPD objectives remain largely…
arXiv:2610.00333v1 Announce Type: cross Abstract: Multimodal on-policy distillation (OPD) aims to improve visual grounding while preserving the strong reasoning capabilities of language models. Recent multi-teacher approaches combine LLM and VLM teachers to provide complementary …
arXiv:2609.38360v1 Announce Type: cross Abstract: On-policy distillation (OPD) has recently emerged as a promising post-training paradigm in which the student learns from trajectories generated by its own policy under dense teacher supervision. However, OPD introduces a fundament…
Multi-teacher on-policy distillation (MOPD) aims to combine the strengths of RL-trained teachers in a single student, but how teacher signals affect parameter changes remains underexplored. We study Qwen3-1.7B with four domain teachers trained with RL from the same initialization…
On-policy distillation (OPD) trains a student on the responses it generates. Existing LLM multi-teacher OPD transfers what specialists predict through their output distributions. We introduce Latent-MOPD, to our knowledge the first representation-level multi-teacher OPD method fo…
arXiv:2609.37510v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student on its own reasoning trajectories using feedback from a stronger teacher. Teacher interventions can improve these trajectories, but also change the distribution on which the student lear…
arXiv:2609.36601v1 Announce Type: new Abstract: On-policy distillation (OPD) reduces train-test state mismatch by training a student on its own generated trajectories, but weak students may visit teacher-misaligned prefixes where supervision is less representative. We introduce S…
Multimodal on-policy distillation (OPD) aims to improve visual grounding while preserving the strong reasoning capabilities of language models. Recent multi-teacher approaches combine LLM and VLM teachers to provide complementary supervision. However, directly using a VLM's full …
arXiv:2609.30837v2 Announce Type: cross Abstract: Multi-teacher on-policy distillation (MOPD) integrates specialized capabilities into a single student, but existing practice typically hard-routes each prompt to a domain-matched teacher for the entire rollout. This dependence on …
On-policy distillation (OPD) has recently emerged as a promising post-training paradigm in which the student learns from trajectories generated by its own policy under dense teacher supervision. However, OPD introduces a fundamental asymmetry: although the sampled trajectories ar…
On-policy distillation (OPD) reduces train-test state mismatch by training a student on its own generated trajectories, but weak students may visit teacher-misaligned prefixes where supervision is less representative. We introduce SAKI (Supervision Allocation with KL-constrained …
Multi-teacher on-policy distillation (MOPD) has emerged as a popular post-training paradigm for integrating specialized capabilities in frontier language models. Existing OPD research has primarily focused on optimizing single-task distillation through objective design, distillat…
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 letti…
Multi-teacher on-policy distillation (MOPD) has emerged as a popular post-training paradigm for integrating specialized capabilities in frontier language models. Existing OPD research has primarily focused on optimizing single-task distillation through objective design, distillat…
On-policy distillation (OPD) trains a student on its own responses with token-level feedback from a stronger teacher, yet the teacher can fail on questions the student already answers correctly, and how often each model succeeds varies across tasks. OPD ignores these outcomes and…
Multi-teacher on-policy distillation (MOPD) integrates specialized capabilities into a single student, but existing practice typically hard-routes each prompt to a domain-matched teacher for the entire rollout. This dependence on prompt-level domain labels restricts using unlabel…
arXiv:2610.08398v1 Announce Type: new Abstract: On-policy distillation from multiple teachers combines expertise from different domains in a single student, but conflicting gradients can hinder this integration. Gradient corrections directly constrain parameter updates under plai…
arXiv:2609.39120v1 Announce Type: new Abstract: On-policy distillation (OPD) improves reasoning by providing token-level supervision from a teacher on a student's own trajectories. Existing methods primarily focus on enhancing this teacher-side guidance (e.g., by enriching teache…