Self-Refining Diffusion Samplers: Enabling Parallelization via Parareal Iterations
–Neural Information Processing Systems
The initial coarse solve (left) via limited steps provides a rough estimate of the sample, which iteratively refined through iterations of our algorithm.
Neural Information Processing Systems
Feb-7-2026, 15:56:14 GMT
- Country:
- Europe
- Germany (0.04)
- Switzerland > Vaud
- Lausanne (0.04)
- North America > United States
- California > Santa Clara County > Palo Alto (0.04)
- Europe
- Genre:
- Research Report > Experimental Study (0.93)
- Technology: