arXiv:2606.27377v1 Announce Type: cross Abstract: Modern image generation demands a single model that unifies diverse capabilities, including text-to-image (T2I), local editing, and global editing. However, these capabilities are rarely naturally aligned and often conflict. For i…
Modern image generation demands a single model that unifies diverse capabilities, including text-to-image (T2I), local editing, and global editing. However, these capabilities are rarely naturally aligned and often conflict. For instance, editing tends to degrade T2I performance,…
arXiv cs.LG
TIER_1English(EN)·Andrei Liviu Nicolicioiu, Mohammad Pezeshki, Aaron Courville·
arXiv:2606.26091v1 Announce Type: new Abstract: On-policy self-distillation achieves strong pass@1 accuracy by using a single model as both teacher and student, with the teacher conditioned on a correct demonstration to provide dense token-level feedback. We show that this could …
A novel on-policy generative field distillation framework called DanceOPD is proposed to unify text-to-image generation, local editing, and global editing capabilities in flow-matching models through capability-specific routing and velocity-based training.
On-policy self-distillation achieves strong pass@1 accuracy by using a single model as both teacher and student, with the teacher conditioned on a correct demonstration to provide dense token-level feedback. We show that this could come at a hidden cost: rollout diversity decreas…
arXiv:2606.22600v2 Announce Type: replace-cross Abstract: On-Policy Distillation (OPD) improves the learning efficiency of standard reinforcement learning through dense, token-level supervision from teachers. In the standard KL objective of OPD, token-level losses are uniformly a…
arXiv cs.LG
TIER_1English(EN)·Wonjun Kang, Kevin Galim, Seunghyuk Oh, Minjun Kang, Sanghyun Park, Donghoon Kim, Minjae Lee, Minseo Kim, Rishabh Tiwari, Yuchen Zeng, Hyung Il Koo, Kangwook Lee·
arXiv:2606.24143v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student on its own rollouts guided by teacher feedback and is becoming increasingly important for large language model (LLM) post-training. Like reinforcement learning (RL), however, OPD faces a…
arXiv:2606.24084v1 Announce Type: cross Abstract: On-policy distillation (OPD) trains a student policy using teacher signals computed on trajectories sampled by the student itself. Recent work shows that sampled-token OPD can be fragile on long-horizon reasoning tasks and that lo…
On-policy distillation (OPD) trains a student on its own rollouts guided by teacher feedback and is becoming increasingly important for large language model (LLM) post-training. Like reinforcement learning (RL), however, OPD faces an on-policy systems bottleneck, as rollouts can …
On-policy distillation (OPD) trains a student policy using teacher signals computed on trajectories sampled by the student itself. Recent work shows that sampled-token OPD can be fragile on long-horizon reasoning tasks and that local teacher-support matching is a simple and effec…
ReNIO enhances on-policy distillation for language models by reweighting negative trajectories based on token-level probability ratios, improving reasoning performance in mathematical and code generation tasks.
On-Policy Distillation (OPD) improves the learning efficiency of standard reinforcement learning through dense, token-level supervision from teachers. In the standard KL objective of OPD, token-level losses are uniformly averaged, implying equal weights for all tokens. However, w…
