Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.06778 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910994745262080 |
|---|---|
| author | Yu, Longlin Zha, Jiajun Yang, Tong Xie, Tianyu Zhang, Xiangyu Chan, S. -H. Gary Zhang, Cheng |
| author_facet | Yu, Longlin Zha, Jiajun Yang, Tong Xie, Tianyu Zhang, Xiangyu Chan, S. -H. Gary Zhang, Cheng |
| contents | Semi-implicit distributions have shown great promise in variational inference and generative modeling. Hierarchical semi-implicit models, which stack multiple semi-implicit layers, enhance the expressiveness of semi-implicit distributions and can be used to accelerate diffusion models given pretrained score networks. However, their sequential training often suffers from slow convergence. In this paper, we introduce CoSIM, a continuous semi-implicit model that extends hierarchical semi-implicit models into a continuous framework. By incorporating a continuous transition kernel, CoSIM enables efficient, simulation-free training. Furthermore, we show that CoSIM achieves consistency with a carefully designed transition kernel, offering a novel approach for multistep distillation of generative models at the distributional level. Extensive experiments on image generation demonstrate that CoSIM performs on par or better than existing diffusion model acceleration methods, achieving superior performance on FD-DINOv2. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_06778 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Continuous Semi-Implicit Models Yu, Longlin Zha, Jiajun Yang, Tong Xie, Tianyu Zhang, Xiangyu Chan, S. -H. Gary Zhang, Cheng Machine Learning Semi-implicit distributions have shown great promise in variational inference and generative modeling. Hierarchical semi-implicit models, which stack multiple semi-implicit layers, enhance the expressiveness of semi-implicit distributions and can be used to accelerate diffusion models given pretrained score networks. However, their sequential training often suffers from slow convergence. In this paper, we introduce CoSIM, a continuous semi-implicit model that extends hierarchical semi-implicit models into a continuous framework. By incorporating a continuous transition kernel, CoSIM enables efficient, simulation-free training. Furthermore, we show that CoSIM achieves consistency with a carefully designed transition kernel, offering a novel approach for multistep distillation of generative models at the distributional level. Extensive experiments on image generation demonstrate that CoSIM performs on par or better than existing diffusion model acceleration methods, achieving superior performance on FD-DINOv2. |
| title | Continuous Semi-Implicit Models |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2506.06778 |