AI Paper Atlas

Research map / Video / Video generation

EFlow: Fast Few-Step Video Generator Training from Scratch via Efficient Solution Flow

Dogyun Park, Yanyu Li, Sergey Tulyakov, Anil Kag

ECCV 2026 · Video › Video generation · approach: Efficient video diffusion models · 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
add Mean-Velocity Additivity regularizer (ablation)EFlow PDG-SFM · Kinetics-700FVD: 185.2 -> 146.7Improved
replace Sana attention with GLGA (ablation)FM VideoDiT · Kinetics-700FVD: 217.7 -> 191.3Improved
replace softmax with GLGA attention (ablation)FM VideoDiT · Kinetics-700FVD: 188.9 -> 191.3No clear effect
add 75% token dropping to Sana attention model (ablation)Sana VideoDiT · Kinetics-700FVD: 217.7 -> 456.1Hurt
add 75% token dropping to GLGA model (ablation)GLGA VideoDiT · Kinetics-700FVD: 191.3 -> 189.7No clear effect
add sliding window attention to linear attention (ablation)GLGA (300K iters) · Kinetics-700FVD: 420.8 -> 352.9Improved
add input-aware gating to summed linear+window attention (ablation)GLGA (300K iters) · Kinetics-700FVD: 352.9 -> 325.7Improved
add linear attention to sliding window attention (ablation)GLGA (300K iters) · Kinetics-700FVD: 390.2 -> 352.9Improved
use PDG-SFM instead of standard SFM at 4 NFE (ablation)EFlow (300K iters) · Kinetics-700FVD: 347.1 -> 322.2Improved
use PDG-SFM at 4 NFE instead of FM at 4 NFE (ablation)EFlow (300K iters) · Kinetics-700FVD: 513.8 -> 322.2Improved
use PDG-SFM 4 NFE instead of FM 40 NFE (GLGA, token drop) (ablation)EFlow · Kinetics-700FVD: 189.7 -> 185.2Improved
add path-drop guidance to cut training cost of CFGEFlow · 480p video training throughputspeedup: 1.76x and 2.5x higher training throughputImproved

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