# When Does a Diffusion Model Decide What to Draw ?

arXiv: https://arxiv.org/abs/2610.05645 · HTML: https://mlsanity.com/paper/2610.05645/

## Key facts

- No citing papers in the map yet (data through 7 Oct 2026).
- First posted on arXiv 5 Oct 2026.

## Placed in

- Diffusion models

## Abstract (arXiv)

A diffusion model starts from pure noise and removes it step by step. Somewhere along the way it stops being able to become "anything" and becomes committed to, say, a horse rather than a truck. We measure when this happens, and what a trained model gets wrong about it, on CIFAR-10. The most direct measurement is to freeze a half-finished image, restart the generation from that point many times, and count how often each class comes out. We call this probability the committor. Measured this way, the model settles coarse questions (vehicle or animal?) at roughly twice the noise level of fine ones (which animal?). A much cheaper measurement, the noise level at which a classifier's opinion about two classes splits into two distinct groups, gets the order of these decisions right (rank correlation 0.73-0.88) but not their exact timing. We then compare pretrained models with their