English
Related papers

Related papers: Fast and Effective On-policy Distillation from Rea…

200 papers

On-policy distillation is pivotal for transferring reasoning capabilities to capacity-constrained models, yet remains prone to instability and negative transfer. We show that on-policy distillation can be interpreted, both theoretically and…

Machine Learning · Computer Science 2026-03-13 Jongwoo Ko , Sara Abdali , Young Jin Kim , Tianyi Chen , Pashmina Cameron

Context distillation enables language models to internalize in-context knowledge into their parameters. In our work, we propose On-Policy Context Distillation (OPCD), a framework that bridges on-policy distillation with context distillation…

Computation and Language · Computer Science 2026-03-24 Tianzhu Ye , Li Dong , Xun Wu , Shaohan Huang , Furu Wei

Distilling reasoning traces from strong large language models into smaller ones is a promising route to improve intelligence in resource-constrained settings. Existing approaches face a fundamental trade-off: offline distillation from…

Computation and Language · Computer Science 2026-05-15 Yumeng Zhang , Zhengbang Yang , Yevin Nikhel Goonatilake , Zhuangdi Zhu

On-policy distillation (OPD) trains a student on its own trajectories under token-level teacher supervision, but existing methods are capped by a single-teacher capability ceiling: when the teacher errs, the student inherits the error. OPD…

Computation and Language · Computer Science 2026-05-05 Jianze Wang , Ying Liu , Jinlong Chen , Xuchun Hu , Qilong Zhang , Yu Cao , Jun Wang , Hua Yang , Yong Xie , Qianglong Chen

On-policy distillation is an efficient alternative to reinforcement learning, offering dense token-level training signals. However, its reliance on a stronger external teacher has driven recent work on on-policy self-distillation, where the…

Machine Learning · Computer Science 2026-05-07 Xin Yu , Liuchen Liao , Yiwen Zhang , Yingchen Yu , Lingzhou Xue , Qinzhen Guo

On-Policy Self-Distillation (OPSD) is a unified learning framework in which a single large language model acts simultaneously as both teacher and student. Unlike conventional knowledge distillation that relies on a separate, often larger…

Human-Computer Interaction · Computer Science 2026-05-22 Fangming Cui , Sunan Li , Jiahong Li

On-policy self-distillation (OPSD) trains a student on its own rollouts using a privileged teacher, but its standard objective weights all generated tokens equally, implicitly treating the privileged teacher target as equally reliable at…

Machine Learning · Computer Science 2026-05-22 Xiaogeng Liu , Xinyan Wang , Yingzi Ma , Yechao Zhang , Chaowei Xiao

On-policy distillation (OPD) trains a student on its own rollouts with token-level teacher supervision. Recent selective OPD methods exploit the non-uniformity of OPD signals by prioritizing high-entropy or high-disagreement tokens. We…

Machine Learning · Computer Science 2026-05-27 Yuanyi Wang , Su Lu , Yanggan Gu , Pengkai Wang , Yifan Yang , Zhaoyi Yan , Congkai Xie , Jianmin Wu , Hongxia Yang

On-policy distillation (OPD) is an effective post-training paradigm for large language models but requires a live teacher server throughout training, resulting in substantial infrastructure overhead. We investigate whether OPD can be…

Machine Learning · Computer Science 2026-05-11 Yecheng Wu , Song Han , Hai Cai

On-policy self-distillation (OPSD) is an emerging LLM post-training paradigm in which the model serves as its own teacher: conditioned on privileged information such as a reference trace or hint, the same policy provides dense token-level…

Machine Learning · Computer Science 2026-05-22 Hongbin Zhang , Chaozheng Wang , Kehai Chen , Youcheng Pan , Yang Xiang , Jinpeng Wang , Min Zhang

On-policy distillation (OPD) has become a popular training paradigm in the LLM community. This paradigm selects a larger model as the teacher to provide dense, fine-grained signals for each sampled trajectory, in contrast to reinforcement…

Machine Learning · Computer Science 2026-04-09 Chenxu Yang , Chuanyu Qin , Qingyi Si , Minghui Chen , Naibin Gu , Dingyu Yao , Zheng Lin , Weiping Wang , Jiaqi Wang , Nan Duan

We introduce Proximal Policy Distillation (PPD), a novel policy distillation method that integrates student-driven distillation and Proximal Policy Optimization (PPO) to increase sample efficiency and to leverage the additional rewards that…

Machine Learning · Computer Science 2026-05-11 Giacomo Spigler

On-Policy Distillation (OPD) has emerged as a dominant post-training paradigm for large language models, especially for reasoning domains. However, OPD remains unstable in practice due to the high gradient variance of its single-sample…

Machine Learning · Computer Science 2026-05-11 Minjae Oh , Sangjun Song , Gyubin Choi , Yunho Choi , Yohan Jo

On-policy reinforcement learning has become the dominant paradigm for reasoning alignment in large language models, yet its sparse, outcome-level rewards make token-level credit assignment notoriously difficult. On-Policy Distillation (OPD)…

Machine Learning · Computer Science 2026-04-14 Binbin Zheng , Xing Ma , Yiheng Liang , Jingqing Ruan , Xiaoliang Fu , Kepeng Lin , Benchang Zhu , Ke Zeng , Xunliang Cai

On-policy distillation (OPD) transfers reasoning behavior by training a student on teacher feedback along student-generated trajectories, but standard full-rollout training ties every update to a costly completion and can over-allocate…

Machine Learning · Computer Science 2026-05-28 Kun Liang , Chenming Tang , Clive Bai , Weijie Liu , Saiyong Yang , Yunfang Wu

Training-time privileged information (PI) can enable language models to succeed on tasks they would otherwise fail, making it a powerful tool for reinforcement learning in hard, long-horizon settings. However, transferring capabilities…

Machine Learning · Computer Science 2026-02-17 Emiliano Penaloza , Dheeraj Vattikonda , Nicolas Gontier , Alexandre Lacoste , Laurent Charlin , Massimo Caccia

Standard knowledge distillation for autoregressive models often suffers from distribution mismatch. While on-policy methods mitigate this by leveraging student-generated outputs, they rely on computationally expensive Reinforcement Learning…

Machine Learning · Computer Science 2026-05-08 Miao Rang , Zhenni Bi , Hang Zhou , Kai Han , Xuechun Wang , An Xiao , Xinghao Chen , Yunhe Wang , Hanting Chen

GRPO-style RLVR trains reasoning models from multiple on-policy attempts per prompt, but typically uses these attempts only through terminal rewards. We show that a mixed group contains a richer process signal: a correct completion is a…

Machine Learning · Computer Science 2026-05-19 Zhiquan Tan , Yinrong Hong

Inference-time harnesses substantially improve large language models on complex reasoning tasks. However, the intrinsic capabilities of the underlying model remain unchanged by the addition of these external workflows. To bridge this gap,…

Computation and Language · Computer Science 2026-05-12 Zhengyang Zhao , Lu Ma , Wentao Zhang

On-policy knowledge distillation (OPD) trains a student on its own rollouts under token-level supervision from a teacher. Not all token positions matter equally, but existing views of token importance are incomplete. We ask a direct…

Machine Learning · Computer Science 2026-05-22 Yuanda Xu , Hejian Sang , Zhengze Zhou , Ran He , Zhipeng Wang , Alborz Geramifard