AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

DiffusionBlocks: Block-wise Neural Network Training via Diffusion Interpretation

Makoto Shing, Masanori Koyama, Takuya Akiba

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use equi-probability instead of uniform partitioning, layers [4,4,4] (ablation)DiffusionBlocks · CIFAR-10FID: 43.53 -> 38.03Improved
use equi-probability instead of uniform partitioning, layers [3,6,3] (ablation)DiffusionBlocks · CIFAR-10FID: 43.59 -> 41.64Improved
use equi-probability instead of uniform partitioning, layers [6,4,2] (ablation)DiffusionBlocks · CIFAR-10FID: 47.49 -> 45.42Improved
use equi-probability instead of uniform partitioning, layers [2,4,6] (ablation)DiffusionBlocks · CIFAR-10FID: 42.37 -> 40.40Improved
use B=2 blocks instead of end-to-end (B=1) (ablation)DiffusionBlocks DiT-L/2 · ImageNetFID: 12.09 -> 9.90Improved
use B=3 blocks instead of end-to-end (B=1) (ablation)DiffusionBlocks DiT-L/2 · ImageNetFID: 12.09 -> 11.11Improved
use B=6 blocks instead of end-to-end (B=1) (ablation)DiffusionBlocks DiT-L/2 · ImageNetFID: 12.09 -> 14.43Hurt
use B=6 blocks instead of B=1 (ablation)DiffusionBlocks DiT-S/2 · CIFAR-10FID: 39.83 -> 53.32Hurt
use B=2 blocks instead of B=1 (ablation)DiffusionBlocks DiT-S/2 · CIFAR-10FID: 39.83 -> 35.47Improved
use B=4 blocks instead of B=2 (ablation)DiffusionBlocks AR · LM1BMAUVE: 0.61 -> 0.67Improved
use B=6 blocks instead of B=4 (ablation)DiffusionBlocks AR · LM1BMAUVE: 0.67 -> 0.62Hurt
apply DiffusionBlocks instead of Forward-ForwardViT · CIFAR-100accuracy: 7.85 -> 59.30Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy