AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Representation Alignment for Diffusion Transformers without External Components

Dengyang Jiang, Mengmeng Wang, Liuzhuozheng Li, Lei Zhang, Haoyu Wang, Wei Wei, Guang Dai, Yanning Zhang, Jingdong Wang

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 SRA self-alignment (3->8)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 33.02 -> 29.10Improved
align layer 3 to deeper teacher layer 8 instead of the same layer 3->3 (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 37.08 -> 29.10Improved
align student layer 6 with teacher layer 10 instead of early layer 3->8 (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 29.10 -> 34.85Hurt
add projection head (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 34.23 -> 29.10Improved
use teacher with lower noise (interval 0.1) instead of equal noise (0.0) (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 31.07 -> 29.55Improved
use fixed interval 0.2 instead of 0.1 (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 29.55 -> 30.70Hurt
vary lambda from 0.1 to 0.4 (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 30.65, 29.10, 29.28, 29.75No clear effect
use EMA alpha 0.9999 instead of student copy alpha 0.0 (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 35.71 -> 29.10Improved
use EMA alpha 0.9999 instead of 0.996 -> 1.0 schedule (ablation)SRA on SiT-B/2 · ImageNet 256 no CFG, 400KFID: 33.17 -> 29.10Improved
replace SiT-XL/2 baseline with SiT + SRA at 800 epochsSRA on SiT-B/2 · ImageNet 256 with CFGFID: 2.06 -> 1.58Improved
use SRA instead of REPA at 800 epochsSRA on SiT-B/2 · ImageNet 256 with CFGFID: 1.42 -> 1.58Hurt
apply SRA to MMDiTSRA on SiT-B/2 · COCO2014FID: 5.86 -> 4.85Improved

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