New Discretization Method Cuts Sampling Cost for Stochastic Diffusion Models
A paper introduces Forward-Value Discretization, which reduces the neural function evaluations needed by stochastic diffusion samplers to match deterministic samplers' quality, without relying on high-order approximations.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-11 05:00 · arXiv stat.ML
Forward-Value Discretization: Faster Stochastic Diffusion Sampling Without High-Order Approximation