AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

SPRINT: Sparse-Dense Residual Fusion for Efficient Diffusion Transformers

Dogyun Park, Moayed Haji-Ali, Yanyu Li, Willi Menapace, Sergey Tulyakov, Hyunwoo J. Kim, Aliaksandr Siarohin, Anil Kag

ICLR 2026 · Image generation › Autoregressive and latent generation · approach: High-resolution diffusion synthesis

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 SPRINT sparse-dense residual fusion with 75% token droppingImproved SiT-XL/2 · ImageNet 256 (SD-VAE, 400K iters, w/o CFG)FID: 12.8 -> 9.30Improved
add SPRINTImproved SiT-XL/2 · ImageNet 256 (SD-VAE, 1M iters, w/ CFG)FID: 2.36 -> 2.29Improved
add Tread or MDTv2 token dropping at 75% drop ratioImproved SiT-XL/2 · ImageNet 256 (SD-VAE, 1M iters, w/ CFG)FID: 2.36 -> 2.82 (Tread), 4.68 (MDTv2)Hurt
add MicroDiT token droppingImproved SiT-XL/2 · ImageNet 256 (SD-VAE, 1M iters, w/ CFG)FID: 2.36 -> 2.53Hurt
add SPRINT to REPASiT-XL/2 REPA · ImageNet 256 (SD-VAE, 400K iters, w/o CFG)FID: 10.0 -> 8.68Improved
add SPRINT to U-ViTU-ViT-XL/2 · ImageNet 256 (SD-VAE, 400K iters, w/o CFG)FID: 12.1 -> 9.20Improved
replace CFG with Path-Drop GuidanceSPRINT · ImageNet 256 system-levelFID: 1.96 -> 1.62Improved
add dense shallow residual path (ablation)SPRINT (SiT-B/2) · ImageNet 256 ablationFID: 85.1 -> 27.5Improved
add structured group-wise token sampling instead of uniform random (ablation)SPRINT (SiT-B/2) · ImageNet 256 ablationFID: 30.1 -> 27.5Improved
raise token drop ratio from 0 to 75% (ablation)SPRINT (SiT-B/2) · ImageNet 256 ablationFID: 54.1 -> 27.5Improved
raise drop ratio from 75% to 87.5% (ablation)SPRINT (SiT-B/2) · ImageNet 256 ablationFID: 27.5 -> 50.2Hurt
shift layers from middle block to encoder/decoder (3-6-3, 5-2-5) (ablation)SPRINT (SiT-B/2) · ImageNet 256 ablationFID: 27.3 -> 29.1 (3-6-3), 49.2 (5-2-5)Hurt

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