titlePhysLoop: Harnessing LLMs' Physics Awareness for Self-Improving Video Generation Models
authorsYuxuan Chen, Xiaojie Li, Hao Chen, Juncheng Ma, Kewei Zhang, Zhongjing Du, Guoqiang Wei, Fangfang Wang, Zhen Dong, Zhiyang Dou, Liming Zhao, Yan Zeng, Jianwei Zhang, Daquan Zhou
date2026-11-12
venuearXiv (preprint, not peer reviewed)
idarXiv:2611.16748
affiliationDAGroup-PKU
upvotes46 (Hugging Face Daily Papers, 2026-11-15)
linksabs pdf hf code
cite
relatedRethinking World-Action Model for Compositional and In-Context Robotic Manipulation
PyRUA-DataFarm: LLM Agent as a Cost-Efficient Teleoperator for Real-Ready Robot Data
Fewer Tokens, Better Action: GPT-6 Astra Robot Agents with 14% Higher Success Rate but 65% Fewer Tokens
DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence
Open-HumanScale: An Open Recipe for Scaling Human and Robot Experience for Embodied Foundation Models
abstractVideo generation models produce visually realistic videos but still struggle with physical dynamics and causal consistency. Learning from generated videos requires fine-grained physical supervision, yet reliable quality judgments, correction pairs, and controlled variations are difficult to obtain. Large language models (LLMs) possess physical knowledge, but struggle to reliably guide video generation models in producing their training videos. We present PhysLoop, a unified harness designed to support physically grounded self-improvement in video generation models. The harness guides and constrains LLMs throughout video generation, selection, and correction, distilling their physical knowledge into physically consistent videos. The framework supports controllable training-data construction and inference-time video repair. By identifying and correcting failures, PhysLoop marks a key milestone toward recursive self-improvement (RSI) in video generation models with limited human intervention. We release 100K videos spanning diverse, realistic indoor and outdoor scenes, along with generation tools and two evaluation datasets: PhysLoop-Gen and PhysLoop-Repair. We evaluate controllable generation and localized repair on these datasets, respectively, and assess physical behavior on the external Physics-IQ Verified benchmark.

bibtex

@misc{chen2026physloop,
  title={PhysLoop: Harnessing LLMs' Physics Awareness for Self-Improving Video Generation Models},
  author={Yuxuan Chen and Xiaojie Li and Hao Chen and Juncheng Ma and Kewei Zhang and Zhongjing Du and Guoqiang Wei and Fangfang Wang and Zhen Dong and Zhiyang Dou and Liming Zhao and Yan Zeng and Jianwei Zhang and Daquan Zhou},
  year={2026},
  eprint={2611.16748},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2611.16748}
}