Perturbated Gradients Updating within Unit Space for Deep Learning

Open paper
Benchmark Model Rank Results
image-classification-on-cifar-10ViT-B/16 (PUGD)#12Percentage correct: 99.13
image-classification-on-cifar-100ViT-B/16 (PUGD)#6Percentage correct: 93.95
image-classification-on-tiny-imagenet-classificationDeiT-B/16 (PUGD)#5Validation Acc: 91.02%
image-classification-on-tiny-imagenet-classificationViT-B/16 (PUGD)#7Validation Acc: 90.74%

Rank counts only results with a code link.