WEIRDO: WEak resIdual Regularized DOob's h-transform diffusion alignment

Excerpt

arXiv:2609.39531v1 Announce Type: cross Abstract: We study the problem of estimating the guidance that steers the distribution learned by a diffusion generative model toward a tilted target $q_0 \propto w\,p_0$ at inference time. Relying on the stochastic optimal control approach, we observe that the exact drift correction is the gradient of the logarithm of Doob's $h$-function, and we study the problem of estimating it from a sample. In the present paper, we assume that the score of the pretrai