About this paper
PhysOPSD uses executable code as privileged physical context to improve the physical plausibility of video generative models through on-policy self-distillation with dual-level privilege and controlled variations.
While modern video generative models achieve remarkable visual fidelity, they remain noticeably brittle when generating physically plausible dynamics and complex interactions. Targeted post-training is crucial, yet supervised fine-tuning suffers from offline state mismatch, and reinforcement learning is hindered by sparse, ill-defined physical reward functions. On-Policy (Self-)Distillation (OPSD) reconciles on-policy rollouts with dense corrective feedback, shifting the core challenge to designing effective privileged context. However, conventional modalities fall short: text prompts are inherently declarative, whereas real reference videos heavily entangle physical laws with visual…
video generative modelsphysically plausible dynamicssupervised fine-tuningoffline state mismatchreinforcement learningphysical reward functionsOn-Policy (Self-)Distillation (OPSD)on-policy rolloutsdense corrective feedbackprivileged context
- Upvotes
- 3
- Authors
- 6
- Published
- Nov 9, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
Trending history
PhysOPSD: Code as Privilege for Physically Plausible Video Generation was first seen on Hugging Face · Daily papers on Nov 11, 2026. Since then it has appeared in 15 checks on 2 different days. Its best position was #16 of 30, reached on Nov 11, 13:28 UTC. It dropped off the list after Nov 12, 07:28 UTC.