New framework unifies image generation capabilities; research tackles distillation challenges
ByPulseAugur Editorial·[12 sources]·
Researchers have introduced DanceOPD, a novel on-policy generative field distillation framework designed to unify diverse image generation capabilities like text-to-image, local editing, and global editing within a single model. This framework addresses the common issue where combining these capabilities can lead to performance degradation. DanceOPD routes samples to specific capability fields and trains using a velocity MSE objective, enabling the composition of expert capabilities while preserving overall generation quality. Additionally, other research explores on-policy distillation techniques, including methods to mitigate output diversity reduction and address position bias in longer reasoning tasks, as well as asynchronous training pipelines to improve efficiency.
AI
IMPACT
Advances in on-policy distillation and generative field frameworks could lead to more versatile and efficient AI models for image generation and reasoning tasks.
RANK_REASON
The cluster consists of multiple arXiv papers detailing novel research in on-policy distillation and generative models.
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…
