AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Dataset Distillation as Pushforward Optimal Quantization

Hong Ye Tan, Emma Slade

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Distillation for few-step generation · 2 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
add k-means cluster-count weights to the training loss on distilled data (ablation)D4M-style clustering distillation · ImageNet-1K classes (distilled latents)Wasserstein-2 distance: -15.7%Improved
add k-means cluster weights to training lossD4M · ImageNet-1K IPC 10 (ResNet-18)top-1 accuracy (%): 27.9 -> 33.1Improved
add cluster weightsD4M · ImageNet-1K IPC 50 (ResNet-18)top-1 accuracy (%): 55.2 -> 56.2Improved
add cluster weightsD4M · ImageNet-1K IPC 200 (ResNet-101)top-1 accuracy (%): 68.1 -> 68.6Improved
add cluster weights, student MobileNet-V2 with ResNet-18 teacherD4M · ImageNet-1K IPC 50 cross-archtop-1 accuracy (%): 47.9 -> 52.1Improved
add cluster weights, Swin-T studentD4M · ImageNet-1K IPC 50 cross-archtop-1 accuracy (%): 58.1 -> 57.4Hurt
use DDOQ clustering instead of RDED patch images at IPC 10RDED · ImageNet-1K IPC 10 (ResNet-18)top-1 accuracy (%): 42.0 -> 33.1Hurt
swap UNet LDM for DiT backbone in DDOQDDOQ · ImageNet-1K IPC 10 (ResNet-18)top-1 accuracy (%): 33.1 -> 53.0Improved
use latent clustering init instead of IGD guidance on DiTMinimax-IGD · ImageNet-1K IPC 10 (DiT backbone)top-1 accuracy (%): 46.2 -> 53.0Improved
use latent clustering init instead of IGD guidance on DiTMinimax-IGD · ImageWoof IPC 10accuracy (%): 43.3 -> 48.8Improved
use latent clustering init instead of IGD guidance on DiTMinimax-IGD · ImageNette IPC 50accuracy (%): 82.3 -> 79.8Hurt

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