xxxnell/how-do-vits-work
(ICLR 2022 Spotlight) Official PyTorch implementation of "How Do Vision Transformers Work?"
PythonApache-2.0
Issues
- 1
How to generate frequency-based noise
#44 opened by JFz0419 - 2
How do I implement Figure2(b) in detail?
#43 opened by pi-wo - 7
How to plot the Hessian max eigenvalue spectra?
#12 opened by Dong1P - 1
- 1
Lesion study
#41 opened by liguopeng0923 - 3
pretrained models
#37 opened by chenrunbin123 - 1
Understanding loss landscape
#40 opened by SonHyegang - 1
relative log magnitude
#39 opened by zhenyuan1234 - 1
AlterNet on CIFAR10
#38 opened by 23Uday - 6
Question about Figure 2(a)
#36 opened by iumyx2612 - 2
Question about harmonizing Convs with MSAs
#35 opened by iumyx2612 - 1
Hessian Max eigenvalue spectra 코드 관련 질문드립니다.
#34 opened by Levinna - 1
Total parameters in AlterNet
#33 opened by sauravtii - 1
pretrained model file is corrupted
#32 opened by youngandbin - 6
- 1
What factors determine if a model or a layer behaves like a low- or high-pass filter?
#31 opened by waitingcheung - 7
What exactly makes MSAs data specificity?
#26 opened by iumyx2612 - 1
- 4
Findings not compatible with other work?
#27 opened by iumyx2612 - 1
ViT vs ResNet: Did you use SAM optimizer?
#28 opened by quannguyen268 - 1
Frequency Analysis for MoCo-v3
#25 opened by YuanLiuuuuuu - 4
- 1
question about detail--drop_pro parameter sd
#24 opened by salt-fisher - 2
- 2
- 1
∆ Log amplitude 관련 질문드립니다.
#21 opened by jhcha08 - 1
Hi,something about object detection...
#20 opened by ross-Hr - 1
Image
#18 opened by 123456789-qwer - 1
- 2
- 1
- 1
- 5
model size
#15 opened by forever10086 - 1
TensorFlow implementation
#14 opened by andreped - 3
Trained models
#8 opened by luluenen - 5
Code for Alter-ResNet-50
#1 opened by DarrenIm - 15
- 3
- 2
헤시안 관련해서 질문드립니다.
#11 opened by Yoontae6719 - 4
how to compute feature map variances?
#6 opened by LostXine - 4
how is robustness calculated?
#5 opened by psteinb - 5
- 0
Plot for
#3 opened by xingchenzhao