AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

Relational Feature Caching for Accelerating Diffusion Transformers

Byunggwan Son, Jeimin Jeon, Jeongwoo Choi, Bumsub Ham

ICLR 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration · 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 relational feature estimation (RFE) to TaylorSeer (ablation)TaylorSeer · ImageNet DiT-XL/2 NFC 11sFID: 5.85 -> 5.22Improved
add relational cache scheduling (RCS) to TaylorSeer (ablation)TaylorSeer · ImageNet DiT-XL/2 NFC 11sFID: 5.85 -> 5.21Improved
combine RFE and RCS (ablation)RFE only · ImageNet DiT-XL/2 NFC 11sFID: 5.22 -> 4.88Improved
replace linear extrapolation with RFE for output change estimate (ablation)caching forecast · ImageNet DiT-XL/2 NFC 14sFID2FC: 3.40 -> 2.02Improved
use RFE instead of fixed scaling w(t)=1.0 (ablation)caching forecast · ImageNet DiT-XL/2 NFC 11FID2FC: 0.73 -> 0.55Improved
schedule by input prediction error (RCS) instead of input distance (ablation)cache scheduling · ImageNet DiT-XL/2 NFC 14sFID2FC: 2.53 -> 1.88Improved
sum input prediction errors over all modules instead of first module only (ablation)RCS · ImageNet DiT-XL/2 NFC 11FID2FC: 0.62 -> 0.61No clear effect
skip full computation more often in shallow layers (ablation)RFC · ImageNet DiT-XL/2 NFC 8 (m=2)FID: 3.40 -> 11.70Hurt
skip full computation more often in deep layers (ablation)RFC · ImageNet DiT-XL/2 NFC 8 (m=2)FID: 3.40 -> 10.20Hurt
use RFC instead of TaylorSeer under large intervalTaylorSeer · ImageNet DiT-XL/2 NFC 7FID: 4.90 -> 3.40Improved
use RFC instead of TaylorSeer on distilled FLUX.1 schnell, 3 NFCTaylorSeer · FLUX.1 schnell DrawBenchPSNR: 22.2134 -> 27.1185Improved
use RFC instead of TaylorSeer on LSUN-bedroom U-Net, NFC 10TaylorSeer · DDIM LSUN-bedroomFID2FC: 37.81 -> 11.64Improved

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