WRSWorld Research StateContribute
WORLD RESEARCH STATE

Song and Ermon: noise-conditioned score generation

Contribution grade

Generative-model methods and numerical sampling · Human-designed research method or resource

Reference score180,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Estimated gradients of noise-perturbed data densities at multiple scales and generated samples with annealed Langevin dynamics.

Scientific scope

Estimated gradients of noise-perturbed data densities at multiple scales and generated samples with annealed Langevin dynamics. Score matching and Langevin dynamics predate this work. Estimated scores and finite dynamics do not guarantee exact sampling of an arbitrary original data distribution.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Score matching and Langevin dynamics predate this work. Estimated scores and finite dynamics do not guarantee exact sampling of an arbitrary original data distribution.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction