AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Adapting Self-Supervised Representations as a Latent Space for Efficient Generation

Ming Gui, Johannes Schusterbauer, Timy Phan, Felix Krause, Joshua M. Susskind, Miguel Ángel Bautista, Björn Ommer

ICLR 2026 · Image generation › Autoregressive and latent generation · approach: Generative latent space tooling · 9 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
use pre-trained SSL encoder (DINOv2) instead of no prior (ablation)RepTok · ImageNet class-cond (10k FID)gFID: 128.54 -> 20.75Improved
use no SSL prior instead of SSL (ablation)RepTok · ImageNet class-cond (10k FID)PSNR: 19.64 vs 14.94Hurt
use DINOv2 instead of CLIP as SSL encoder (ablation)RepTok · ImageNet (10k FID)gFID: 30.56 -> 20.75Improved
use DINOv2 instead of MAE as SSL encoder (ablation)RepTok · ImageNet (10k FID)gFID: 28.48 -> 20.75Improved
use DINOv2 instead of CLIP (ablation)RepTok · ImageNet (10k FID)rFID: 13.66 -> 7.95Improved
use single-token latent with MLP-Mixer instead of SiT-XL/2 latent diffusionRepTok · ImageNet training computeTotal PFlops: 212.5K -> 34.5KImproved
add CFG=1.5RepTok · ImageNetgFID: 3.4 -> 3.22Improved
scale generator to RepTok-L (516M params)RepTok · ImageNetgFID: 3.4 -> 2.06Improved
fine-tune cls token instead of freezing it (ablation)RepTok · ImageNet reconstructionreconstruction qualityImproved
increase cosine-similarity weight lambda (ablation)RepTok · ImageNetgFID vs PSNRMixed
use DINOv2 register tokens instead of cls token (frozen) (ablation)RepTok decoder · ImageNet reconstructionPSNR / rFID: PSNR 12.59 -> 12.85; rFID 14.13 -> 14.90Mixed
add register tokens as extra latent tokens (ablation)RepTok · ImageNetgFIDHurt

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