issues
search
jpzhang1810
/
TGR
Official Pytorch implementation for "Transferable Adversarial Attacks on Vision Transformers with Token Gradient Regularization" (CVPR 2023).
25
stars
5
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
关于ViTs的权重问题
#8
MisterRpeng
closed
6 months ago
0
参数无法取出
#7
Evalution13
opened
8 months ago
2
v通道获取问题
#6
dazhonghua119
opened
8 months ago
0
你好,因为代码兼容性问题,我将register_backward_hook修改成了register_full_forward_hook,并对v_tgr的返回进行了一些修改,可是最后执行结果有差异。 另外我在启用PatchOut之后攻击成功率反而降低了,感觉奇怪。这是我的v_tgr的返回,以及我的环境和执行结果:
#5
guox18
closed
1 year ago
0
你好,因为代码兼容性问题,我将register_backward_hook修改成了register_full_forward_hook,并对v_tgr的返回进行了一些修改,可是最后执行结果有差异。这是我的v_tgr的返回,以及我的环境和执行结果:
#4
guox18
closed
1 year ago
0
关于token gradients概念的不理解
#3
WateverOk
opened
1 year ago
2
按照文档配置环境,但出现了如下的错误
#2
UsernameRance
closed
1 year ago
3
Gradient dimension mismatch
#1
zeyuanyin
closed
1 year ago
2