AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

DiCache: Let Diffusion Model Determine Its Own Cache

Jiazi Bu, Pengyang Ling, Yujie Zhou, Yibin Wang, Yuhang Zang, Dahua Lin, Jiaqi Wang

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration

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 Dynamic Cache Trajectory Alignment (DCTA) (ablation)DiCache · HunyuanVideoPSNR: 31.98 -> 32.79Improved
add DCTA (SSIM/LPIPS) (ablation)DiCache · HunyuanVideoSSIM: 0.9314 -> 0.9396Improved
increase probe depth from m=1 to m=5 (ablation)DiCache online probe · HunyuanVideoLPIPS and speedup: LPIPS 0.1492 -> 0.1367; speedup 2.34x -> 2.10xMixed
lower reuse threshold from 0.10 to 0.05 (ablation)DiCache · HunyuanVideoLPIPS and speedup: LPIPS 0.1492 -> 0.1047; speedup 2.34x -> 1.76xMixed
raise reuse threshold from 0.10 to 0.20 (ablation)DiCache · HunyuanVideoLPIPS and speedup: LPIPS 0.1492 -> 0.1886; speedup 2.34x -> 2.90xMixed
use shallow-layer probe (online) instead of input differences as caching error indicatorDiCache · Wan/Hunyuan/Flux feature analysisSpearman correlation with output differences: around 0.8 at m in [1,3]Improved
combine DiCache with Sparse VideoGenHunyuanVideo · HunyuanVideo 720x1280, 129 framesspeedup: 1.67x -> 3.08xImproved
combine DiCache with Sparse VideoGenHunyuanVideo · HunyuanVideo 720x1280, 129 framesLPIPS: 0.1760 -> 0.2051Hurt
use offline-prior or empirical-law caching (TeaCache-fast) instead of online probeTeaCache · HunyuanVideoLPIPS: 0.1492 -> 0.2898Hurt

Later papers that compared with this method

3 later accepted papers report DiCache as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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