AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

TC-Pade: Trajectory-Consistent Pade Approximation for Diffusion Acceleration

Shaoxuan He, Benlei Cui, Bukun Huang, Zhizeng Ye, Yunyun Sun, Longtao Huang, Hui Xue, Yang Yang, Haiwen Hong, Jingqun Tang, Zhou Zhao

CVPR 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
cache residuals across the entire block instead of double-stream blocks only (ablation)TC-Padé · DrawBench, FLUX.1-devImageReward: 0.7921 -> 0.9184Improved
cache residuals across the entire block instead of single-stream blocks only (ablation)TC-Padé · DrawBench, FLUX.1-devImageReward: 0.8717 -> 0.9184Improved
cache residuals across the entire block instead of double-stream only (ablation)TC-Padé · DrawBench, FLUX.1-devlatency speedup: 1.36× -> 2.88×Improved
lower TSI threshold from 1.3 to 0.7 (more aggressive skipping) (ablation)TC-Padé · DrawBench, FLUX.1-devspeedup: 1.63× -> 2.88×Improved
lower TSI threshold from 1.3 to 0.7 (more aggressive skipping) (ablation)TC-Padé · DrawBench, FLUX.1-devImageReward: 0.9562 -> 0.9184Hurt
predict residuals with Padé instead of raw features with Taylor series (TC-Padé vs TaylorSeer)TC-Padé · COCO 2017, FLUX.1-dev, 20 stepsPSNR: 16.567 -> 21.962Improved
predict residuals with Padé instead of Taylor extrapolation on videoTC-Padé · Wan2.1-1.3B, VBench-2.0VBench-2.0 score (%): 54.50% -> 60.38%Improved
apply TC-Padé to class-conditional DiT-XL/2DiT-XL/2 · ImageNet 256x256, 20 stepsFID: 3.56 -> 6.93Hurt
apply TC-Padé acceleration to FLUX.1-devFLUX.1-dev · COCO 2017, 20 stepsFID: 23.38 -> 24.14Hurt
add quantization on top of TC-PadéTC-Padé on FLUX.1-dev · COCO 2017, 20 stepsCLIP score: 31.82 -> 31.08Hurt
add quantization on top of TC-PadéTC-Padé on FLUX.1-dev · COCO 2017, 20 stepsFID: 24.14 -> 24.31No clear effect

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