SDE Matching Extends to Continuous Physical Time, Advancing Simulation-free Latent SDEs
Recent works extend SDE Matching from one-shot marginal distributions to learning generative models directly in continuous physical time, bypassing the pseudo-time limitations of diffusion models and flow matching. A related paper closes the approximation gap in simulation-free latent SDEs.
2026-10-06 ~ 2026-10-06 · 2 related posts
- Helmholtz-SDE closes the approximation gap in simulation-free latent SDE inference — canaesseth · 2026-10-06
- SDE Matching extended beyond one-time marginals to learn generative models in physical time — canaesseth · 2026-10-06