GigaTok-XL-XXL

Reported on 4 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical4 results

Image ReconstructiononImageNet
FID· 2025-04-11
0.79
best: 0.49 (MGVQ (16x16x8))
SOTA
GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation arXiv:2504.08736
Image ReconstructiononImageNet
LPIPS· 2025-04-11
0.1947
best: 0.066 (OptVQ (16x16x8))
GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation arXiv:2504.08736
Image ReconstructiononImageNet
PSNR· 2025-04-11
21.65
best: 27.57 (OptVQ (16x16x8))
GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation arXiv:2504.08736
Image ReconstructiononImageNet
SSIM· 2025-04-11
0.699
best: 0.787 (MGVQ (16x16x8))
GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation arXiv:2504.08736