Paper page - On-Policy Adversarial Flow Distillation for Autoregressive Video Generation

Source: https://huggingface.co/papers/2605.26105

Abstract

Adversarial Flow Distillation enables efficient distillation of heterogeneous video generation models by using on-policy feedback and forward-process flow-matching updates without requiring teacher scores or detailed trajectory information.

Autoregressive video generatorsare attractive for streaming, long-horizon, and interactive applications, but distilling strongblack-box teachersintocausal studentsremains difficult. The student must learn under its own rollout distribution, whereas practical teachers may expose only prompt-conditioned completed videos and may differ in architecture, capacity, temporal design, and sampling schedule. This interface makes supervised fine-tuning off-policy, score-based distillation inapplicable, and direct adversarial imitation too sparse for denoising-time credit assignment. We propose Adversarial Flow Distillation (AFD), anon-policy frameworkfor heterogeneous black-boxvideo distillation. AFD queries the teacher and rolls out the current student on the same prompts, trains a prompt-pairedBradley-Terry discriminatorto estimate clean-sample teacher-student discrepancy, and converts the resulting on-policy advantage into forward-processflow-matching updateson the student’s own noised states. Thus, AFD provides dense velocity-field supervision while requiring no teacher scores, latents, denoising trajectories, step alignment, or reverse-chain reinforcement learning. Experiments across two causal AR student families show that AFD consistently improves motion- and physics-sensitive generation while preserving general video quality, and ablations validate the importance of adaptive on-policy feedback andforward-process credit assignment. The method requires only clean teacher videos and student rollouts, providing a practical route for distilling proprietary or heterogeneous video generators into efficient autoregressive students.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.26105

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.26105 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.26105 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.26105 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.