Issues
- 1
Support tabular data?
#102 opened by Hiramdu - 12
ImportError: cannot import name 'zero_gradients' from 'torch.autograd.gradcheck'
#99 opened by ashok-arjun - 3
weird FGSM accuracy on MNIST clean data
#111 opened by chhyun - 0
Implementation of CW Attack in Linf distance
#114 opened by youyuanyi - 2
Is it expected that for gradient-based methods, the parameter gradients in the `prediction()` may accumulate?
#113 opened by LorrinWWW - 1
Gradient calculation issue about PGD attacks
#85 opened by rshaojimmy - 0
How long does it take to generate adversarial examples when PGD-Adversarial Training with CIFAR10?
#109 opened by chhyun - 0
- 1
L2 Momentum Iterative Attack, risk of NA outputs
#105 opened by qlero - 1
Robust model weights
#106 opened by AvivSham - 1
CarliniWagnerL2Attack on MNIST does NOT work
#101 opened by wenhaoyong - 4
- 3
- 0
L1PGDAttack switches cuda devices
#96 opened by siebenkopf - 3
- 2
name 'adversary' is not defined
#90 opened by YOONAHLEE - 0
Is batch attack supported for PGD?
#92 opened by minkyu-choi04 - 1
- 1
Attack on ensemble models?
#88 opened by taoye9 - 1
multi_gpu
#87 opened by Fengww2014 - 0
- 0
adversary.perturb(self, x, y)
#84 opened by SoEzreal - 4
JSMA never success on CIFAR10
#31 opened by amosy3 - 3
Adversarial training on cifar10?
#76 opened by Originofamonia - 3
CarliniWagnerL2Attack takes forever on CIFAR10
#83 opened by dvd42 - 4
How does PGDAttack deal with embedding?
#81 opened by XueAdas - 3
add simple spatial transform attack
#80 opened by MasanoriYamada - 2
- 8
L1 PGD attack issue
#48 opened by JungeumKim - 5
- 1
- 1
Why cw_loss has a 50 intercept?
#72 opened by zzzace2000 - 2
Some bugs on LocalSearchAttack
#71 opened by 543877815 - 6
- 3
- 11
- 2
- 4
- 2
In tutorial_train_mnist.py, what is the reason for choosing 'mean' or 'sum' in F.cross_entropy
#56 opened by TianQi-777 - 3
Targeted attacks increases accuracy to 100%
#55 opened by awaisrauf - 4
Failed adv training: MNIST and 2_layer_MLP
#54 opened by TianQi-777 - 1
- 2
FGSM return tensor cannot compute the gradient
#49 opened by xunge - 2
Question between two similar attacks
#43 opened by averyma - 3
What if predict(network) has multiple outputs?
#33 opened by CandiceUIC - 5
question for iterative l2 attack
#39 opened by ZekunStevenZhang - 2
Error loading notebook
#36 opened by nocotan - 4
Please add deepfool/DDN support
#25 opened by drmeerkat - 4
Printing attack adversaries in a verbose mode, for better debugging purpose
#32 opened by stefan-matcovici - 4
Attack successfully all the time?
#29 opened by machengcheng2016