Diffusion-Informed Single-pass Classifier

Classifiers built from diffusion models (i.e. diffusion classifiers) are reported to be robust and well aligned with human perception. However, a single prediction requires about $10^5$ forward passes, which is far too costly for deployment. We ask where this alignment (shape bias, out-of-distribution accuracy, and error consistency with human mistakes) comes from, and whether an ordinary single-pass discriminative classifier can have it too. We carefully decompose the human alignment gap between diffusion classifiers and discriminative classifiers that is reported in the literature into the axes along which their designs differ: training data, resolution, aggregation of evidence across noise levels and training objective (generative vs discriminative). We isolate “aggregation” as an important contributor (more than training data or the generative objective alone) to this alignment. We measure how much each noise level contributes to a diffusion classifier's prediction through reductions in KL divergence to its final prediction. The normalized average contributions define a noise-sampling distribution, $w_{\mathrm{gen}}$, and their cumulative sum, $\rho$, provides a label-smoothing schedule. We use these two curves to set the noise levels and label confidence when training an ordinary single-pass classifier (e.g., a ResNet-50), transferring aggregation with no change to architecture or inference cost; and, unlike naive noise augmentation, at no cost in accuracy. Our recipe, DISC (Diffusion-informed single-pass classifier), matches the diffusion classifier's out-of-distribution accuracy, exceeds its error consistency, narrows its shape-bias gap, and improves robustness to corruptions and distribution shifts without losing clean accuracy. The $w_{\mathrm{gen}}$ curve also serves as an adaptive guidance schedule, improving diffusion sampling quality.