AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Cubic Discrete Diffusion: Discrete Visual Generation on High-Dimensional Representation Tokens

Yuqing Wang, Chuofan Ma, Zhijie Lin, Yao Teng, Lijun Yu, Shuai Wang, Jiaming Han, Jiashi Feng, Yi Jiang, Xihui Liu

CVPR 2026 · Image generation › Autoregressive and latent generation · approach: Latent diffusion with learned tokenizers · 0 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
mask individual elements across the tensor instead of whole spatial positions (ablation)CubiD-L · ImageNet 256x256gFID: 22.22 -> 5.33Improved
mask individual elements instead of whole dimensions across positions (ablation)CubiD-L · ImageNet 256x256gFID: 120.03 -> 5.33Improved
use a learned mask token instead of a random codebook token (ablation)CubiD-L · ImageNet 256x256gFID: 56.38 -> 5.33Improved
use a learned mask token instead of a fixed zero token (ablation)CubiD-L · ImageNet 256x256gFID: 5.56 -> 5.33Improved
use mask ratio std sigma=0.10 rather than 0.05 (ablation)CubiD-L · ImageNet 256x256gFID: 7.65 -> 5.33Improved
use mask ratio std sigma=0.10 rather than 0.15 (ablation)CubiD-L · ImageNet 256x256gFID: 5.81 -> 5.33Improved
increase inference steps from 64 to 256 (ablation)CubiD-L · ImageNet 256x256gFID: 9.14 -> 5.33Improved
increase inference steps from 256 to 512 (ablation)CubiD-L · ImageNet 256x256gFID: 5.33 -> 5.25No clear effect
scale model from 946M to 3.7B parameters (ablation)CubiD · ImageNet 256x256gFID: 5.25 -> 4.68Improved
use SigLIP2 instead of DINOv2 as encoder (ablation)CubiD-L · ImageNet 256x256gFID: 5.25 -> 5.87Hurt
use dimension-wise quantization instead of vector quantization on SigLIP2 features (ablation)LLaVA · GQAaccuracy: 54.9 -> 63.1Improved
increase quantization levels L from 2 to 8 (ablation)DINOv2-B dimension-wise quantization · ImageNet reconstructionrFID: 1.38 -> 0.57Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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