List your company

Image Generation on ImageNet 256x256

478.6ISProMoE-XL-Flow
292.856341.078389.3437.522Dec 4, 2023May 8, 2024Oct 12, 2024Mar 18, 2025Aug 22, 2025Jan 26, 2026Jul 2, 2026

Evaluation results

MethodLinks
ProMoE-XL-FlowTraining Steps=2M, # Activated Params.=675M, # Total Params.=1.568B, cfg=4.02025.10478.6————————11.95—————
DiffMoE-XL-FlowTraining Steps=2M, # Activated Params.=675M, # Total Params.=1.735B, cfg=4.02025.10475.13————————14.25—————
Dense-DiT-XL-FlowTraining Steps=2M, # Activated Params.=675M, # Total Params.=675M, cfg=4.02025.10474.39————————17.05—————
ProMoE-L-FlowTraining Steps=2M, # Activated Params.=458M, # Total Params.=1.063B, cfg=4.02025.10466.96————————14.05—————
DiffMoE-L-FlowTraining Steps=2M, # Activated Params.=458M, # Total Params.=1.163B, cfg=4.02025.10465.6————————15.7—————
Dense-DiT-L-FlowTraining Steps=2M, # Activated Params.=458M, # Total Params.=458M, cfg=4.02025.10464.25————————17.71—————
VARTokenizer=VQVAE, Training Epochs=350, #params=2.0B, Classifier-Free Guidance=true2025.04365.41.8—8357——————————
VAREpochs=350, Classifier-Free Guidance (CFG)=true, Category=Autoregressive Models2025.12365.4——83571.8—————————
VAREpochs=350, with CFG=true2026.06365.4——83571.8—————————
MaskGITClass=Masked Modeling2023.12355.6——83444.02—————————
Predictor–CorrectorBackbone=SoFlow-XL/2-cond, K=20, t_noise=0.52026.05349.84—5.3387504.67—————————
FRACTALMAR-HPARAMS=848M, STEPS=962026.02348.96.15—————————————
FractalMAR-HPre-training Tokenizer=-, Pre-training SSL Encoder=-, #Params=848M, #Epochs=6002026.03348.9————6.15—————————
FractalMAR-HParams=848M, Epochs=8002026.05348.9——81466.15—————————
FractalMAR-HParams=848M, Epochs=600, Diffusion Space=Pixel-space2026.06348.9——81466.15—————————
RQTransGenerator Type=AR, Generator #Params=1.4B, Classifier Guidance=True2026.05337.53.89—————————————
AliTok-LGenerator Type=AR, Generator #Params=318M, Classifier Guidance=True2026.05326.21.38—————————————
RQ-TransformerClass=Autoregressive2023.12323.7————3.8—————————
RQ-Tran.Type=AR, Params=3.8B2026.03323.73.8—————————————
VAREpochs=350, #Params=2B, Generation Resolution=256x256, Guidance Setting=w/o guidance2026.01323.11.92—8259——————————
RandAR-XXLType=AR, Params=1.4B2026.033222.15—7962——————————
REGEpochs=80, Classifier-Free Guidance (CFG)=true, Backbone=SiT-XL/2, Category=Leveraging VFM's Representations2025.12321.4—4.4976631.86—————————
REGEpochs=80, with CFG=true2026.06321.4—4.4976631.86—————————
SDE-DPM-Solver++ (2M)Schedule=Time-uniform, NFE=202026.01320.83—7——7.65—————————
REGLUEEpochs=160, Classifier-Free Guidance (CFG)=true, Backbone=SiT-XL/2, Category=Leveraging VFM's Representations2025.12320.4—4.2678651.53—————————
SDE-DPM-Solver++ (2M)Schedule=LAS, NFE=202026.01320.18—6.18——6.67—————————
MagViT-v2Tokenizer=LFQ tokenizers, Training Epochs=1080, #params=307M, Classifier-Free Guidance=true2025.04319.41.78—————————————
MagViTv2Epochs=1080, Classifier-Free Guidance (CFG)=true, Category=Autoregressive Models2025.12319.4——83571.78—————————
MAGVIT-v2Generator Type=Mask, Generator #Params=307M, Classifier Guidance=True2026.05319.41.78—————————————
MagViTv2Epochs=1080, with CFG=true2026.06319.4——83571.78—————————
Checkerboard-LScale Ratio=2x, Classifier-free guidance (cfg)=1.5, Params=343M, #Steps=17, Time (s)=0.522026.02318.2——82572.83—————————
RandAR-XLType=AR, Params=775M2026.03317.82.25—8060——————————
MDTV2Epochs=1080, with CFG=true2026.06317.7—4.5279651.58—————————
SiT-XL/2 + AHPACategory=Latent diffusion, Transformer, Epochs=800, CFG=true, Standalone (w/o Ext.)=true2026.05317.5—4.1281641.4—————————
SiT-XL/2 + SRA 2Category=Latent diffusion, Transformer, Epochs=800, CFG=true, Standalone (w/o Ext.)=true2026.05316.2—4.6382621.52—————————
SiT-XL/2 + oursTraining Steps=4M, Classifier-free Guidance (CFG)=Enabled2026.07315.3—4.681631.44—————————
MDTv2Tokenizer=SD-VAE, Training Epochs=1080, #params=675M, Classifier-Free Guidance=true2025.04314.71.584.527965——————————
MDTv2-XL/2Category=Transformer + U-Net hybrid, Epochs=1080, CFG=True2026.01314.7—4.5279651.58—————————
REGLUEEpochs=80, Classifier-Free Guidance (CFG)=true, Backbone=SiT-XL/2, Category=Leveraging VFM's Representations2025.12313.1—4.2578641.61—————————
VAR-d24Type=AR, Params=1.0B2026.03312.92.09—8259——————————
MaskAlignEpochs=800, with CFG=true2026.06312.9—4.3178671.35—————————
Checkerboard-LScale Ratio=4x, Classifier-free guidance (cfg)=1.7, Params=343M, #Steps=17, Time (s)=0.522026.02311.5——80572.79—————————
SiT-XL/2 + SRACategory=Training Acceleration, Epochs=800, CFG=True2026.01311.4—4.6580631.58—————————
SiT-XL/2 + REPACategory=Latent diffusion, Transformer, Epochs=800, CFG=true, Standalone (w/o Ext.)=false2026.05311.4—4.780631.42—————————
SiT-XL/2 + SRATraining Steps=4M, Classifier-free Guidance (CFG)=Enabled2026.07311.4—4.6580631.58—————————
Self-Transcendence (SiT-XL/2)Category=Training Acceleration, Backbone=SiT-XL/2, Epochs=400, CFG=True2026.01311.3—4.8579661.44—————————
DDT-XL/2Pre-training Tokenizer=SD-VAE, Pre-training SSL Encoder=DINOv2, #Params=675+49M, #Epochs=4002026.03310.6————1.26—————————
DDT-XL/2Framework Category=Two-stage Frameworks with Aux Supervision (DINOv2), Aux. Token=SD-VAE, Params=675M+49M2026.03310.6————1.26—————————
DDT-XL/2Params=675+49M, Epochs=4002026.05310.6——78671.26—————————
DDT-XLParams=675M, Epochs=400, GFLOPs=119, Diffusion Space=Latent-space2026.06310.6——79651.26—————————
SemTok-AR-XXLType=AR, Params=1.2B2026.03310.52.34—7763——————————
MaskAlignEpochs=80, with CFG=true2026.06310—4.4881631.82—————————
UNITE-XLFramework Category=Single-stage Frameworks, Aux. Token=Joint, Params=806M2026.03309.9————1.75—————————
SDE-DPM-Solver++ (2M)Schedule=LogSNR, NFE=202026.01308.36—8.56——7.15—————————
SiT-XL/2 + AHPACategory=Latent diffusion, Transformer, Epochs=400, CFG=true, Standalone (w/o Ext.)=true2026.05308.2—4.1581631.56—————————
PixelU-H/16Params=1168M, Epochs=320, NFE=100×2, GFLOPs=137, Diffusion Space=Pixel-space2026.06306.86—5.2978631.77—————————
HyperDiT-HParams=952M, Epochs=600, NFE=100x22026.05306.5—4.7380641.56—————————
dMALABackbone=SoFlow-XL/2-cond, K=20, σ=0.32026.05306.25—5.8180622.17—————————
PixelU-H/16Params=1168M, Epochs=600, NFE=100×2, GFLOPs=137, Diffusion Space=Pixel-space2026.06305.88—5.0479641.63—————————
REPATokenizer=SD-VAE, Training Epochs=800, #params=675M, Classifier-Free Guidance=true2025.04305.71.284.687964——————————
SiT-XL/2 + REPANFE=250×2, space=latent, params=675M (49M), Gflops=119 (310)2026.01305.7————1.42—————————
SiT-XL/2 + REPACategory=Training Acceleration, Epochs=800, CFG=True2026.01305.7—4.780651.42—————————
REPAEpochs=800, Classifier-Free Guidance (CFG)=true, Backbone=SiT-XL/2, Category=Leveraging VFM's Representations2025.12305.7—4.780651.42—————————
REPA-XLTokenizer=SD-VAE, Training Epochs=800, Steps=250, #params=675M, Classifier-Free Guidance (CFG)=true2025.10305.71.42—————————————
REPA, SiT-XL/2Pre-training Tokenizer=SD-VAE, Pre-training SSL Encoder=DINOv2, #Params=675+49M, #Epochs=8002026.03305.7————1.42—————————
REPA-SiT-XL/2Framework Category=Two-stage Frameworks with Aux Supervision (DINOv2), Aux. Token=SD-VAE, Params=675M+49M2026.03305.7————1.42—————————
REPA-XL/2Generator Type=Diff., Generator #Params=675M, Classifier Guidance=True2026.05305.71.42—————————————
SiT-XL/2 + SRACategory=Latent diffusion, Transformer, Epochs=800, CFG=true, Standalone (w/o Ext.)=true2026.05305.7—4.6580631.58—————————
REPA-XL/2Params=675+49M, Epochs=800, NFE=250x22026.05305.7—4.780641.42—————————
SiT-XL/2+REPA# Params.=675M, NFE=250 × 22026.05305.7————1.42—————————
REPAEpochs=800, with CFG=true2026.06305.7—4.780651.42—————————
REPA-XL/2Params=675M + 86M, Epochs=800, NFE=250×2, GFLOPs=119, Diffusion Space=Latent-space2026.06305.7—4.780641.42—————————
SiT-XL/2 + REPATraining Steps=4M, Classifier-free Guidance (CFG)=Enabled2026.07305.7—4.780651.42—————————
SemTok-AR-XLType=AR, Params=746M2026.03305.62.54—7860——————————
SpectualAR-d20Generator Type=AR, Generator #Params=600M, Classifier Guidance=True2026.05305.42.49—————————————
SiT-XL/2 + Self-FlowTraining Steps=4M, Classifier-free Guidance (CFG)=Enabled2026.07305.4—4.5480611.47—————————
REGEpochs=160, Classifier-Free Guidance (CFG)=true, Backbone=SiT-XL/2, Category=Leveraging VFM's Representations2025.12304.6—4.3677651.59—————————
REGEpochs=160, with CFG=true2026.06304.6—4.3677651.59—————————
ProMoE-XL-FlowTraining Steps=2M, # Activated Params.=675M, # Total Params.=1.568B, cfg=1.52025.10304.26————————2.08—————
HyperDiT-XLParams=676M, Epochs=600, NFE=100x22026.05304.2—4.7680621.63—————————
MaskAlignEpochs=160, with CFG=true2026.06304.1—4.3779651.56—————————
DeCo-XL/16Pre-training Tokenizer=-, Pre-training SSL Encoder=DINOv2, #Params=682M, #Epochs=6002026.03304————1.69—————————
DeCo-XL/16Params=682M, Epochs=600, NFE=100x22026.05304—4.5979631.69—————————
DeCo-XL/16Params=682M, Epochs=600, NFE=100×2, Diffusion Space=Pixel-space2026.06304—4.5979631.69—————————
Self-Transcendence (LightningDiT-XL/1)Category=Training Acceleration, Backbone=LightningDiT-XL/1, Epochs=400, CFG=True2026.01303.9—4.1178661.25—————————
MARTokenizer=LDM, Training Epochs=800, #params=945M, Classifier-Free Guidance=true2025.04303.71.55—8162——————————
MAREpochs=800, Classifier-Free Guidance (CFG)=true, Category=Autoregressive Models2025.12303.7——81621.55—————————
MAREpochs=800, with CFG=true2026.06303.7——81621.55—————————
JIT-HPARAMS=953M, STEPS=1002026.02303.41.86—————————————
JiT-H/16Pre-training Tokenizer=-, Pre-training SSL Encoder=-, #Params=953M, #Epochs=6002026.03303.4————1.86—————————
JiT-H/16Framework Category=Single-stage Frameworks, Aux. Token=-, Params=953M2026.03303.4————1.86—————————
REPATokenizer=E2E-VAE (Ours), Training Epochs=800, #params=675M, Classifier-Free Guidance=true2025.04302.91.124.097966——————————
VAR-d20Type=Parallelized AR, Params.=600M2026.01302.6——83562.95—————————
VAR-d20Type=AR, Params=600M2026.03302.62.57—8356——————————
VAR-d20Generator Type=VAR, Generator #Params=310M, Classifier Guidance=True2026.05302.62.57—————————————
Checkerboard-LScale Ratio=2x, Classifier-free guidance (cfg)=1.4, Params=343M, #Steps=17, Time (s)=0.522026.02302.5——81562.72—————————
SPHERE-LPARAMS=950M, STEPS=42026.02301.84.76—————————————
dMALABackbone=SoFlow-XL/2-cond, K=10, σ=0.32026.05300.04—5.0579651.7—————————
PixNerd-XL/16Params=700M, Epochs=320, NFE=100x22026.05300—4.5480601.95—————————
PixNerd-XL/16Params=700M, Epochs=320, NFE=100×2, GFLOPs=134, Diffusion Space=Pixel-space2026.06300—4.5480601.95—————————