AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Structural Assessment for Understanding and Guiding Dataset Distillation in Discrete Token Space

Yue Cao, Jianyang Gu, Vyacheslav Kungurtsev, Yu Hu, Jozsef Hamari, Zheng Liu, Mohsen Zardadi

ECCV 2026 · Label-efficient and robust learning › Knowledge distillation · approach: Rethinking distillation objectives · 1 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
cluster in discrete token space (no PCA, no anchors) (ablation)TGDD · ImageWoof IPC 20accuracy: 43.6 -> 44.3Improved
add PCA on token histograms (ablation)TGDD · ImageWoof IPC 50accuracy: 57.4 -> 58.9Improved
add structural-score anchor selection (ablation)TGDD · ImageWoof IPC 50accuracy: 58.9 -> 60.3Improved
replace score-based anchor selection with random sampling (ablation)TGDD · ImageWoof IPC 20accuracy: 46.3 -> 43.1Hurt
use all three structural-score terms instead of JSD only (ablation)TGDD · ImageWoof IPC 10accuracy: 38.2 -> 41.2Improved
use discrete tokenizer guidance instead of DINOv2 (ablation)TGDD · ImageWoof IPC 50 (DiT)accuracy: 58.1 -> 60.3Improved
use VQ-VAE instead of VQGAN tokenizer on LDM (ablation)TGDD · ImageWoof (LDM)accuracy: 55.8 -> 55.1Hurt
weight low-resolution scales more ([3,1,0.5]) vs uniform (ablation)TGDD · ImageWoofaccuracy: 58.8 -> 60.3Improved
increase anchors from 10 to 20 at IPC 50 (ablation)TGDD · ImageNette IPC 50accuracy: 81.3 -> 79.9Hurt
TGDD vs MGD3 on ImageNet-1k IPC 50TGDD · ImageNet-1kaccuracy: 60.2 -> 60.3No clear effect
TGDD vs MGD3 on ImageWoof IPC 50TGDD · ImageWoof ResNet-18accuracy: 58.3 -> 61.5Improved
lower token HHI (more balanced token use) in distilled setTGDD · ImageWoof surrogatesaccuracy: w_HHI=-10.99Improved

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