Papers › Explaining and Harnessing Adversarial Examples

Explaining and Harnessing Adversarial Examples

20 Dec 2014arXiv:1412.6572archive 2025-07-28

Ian J. Goodfellow, Jonathon Shlens, Christian Szegedy

Several machine learning models, including neural networks, consistently misclassify adversarial examples---inputs formed by applying small but intentionally worst-case perturbations to examples from the dataset, such that the perturbed input results in the model outputting an incorrect answer with high confidence. Early attempts at explaining this phenomenon focused on nonlinearity and overfitting. We argue instead that the primary cause of neural networks' vulnerability to adversarial perturbation is their linear nature. This explanation is supported by new quantitative results while giving the first explanation of the most intriguing fact about them: their generalization across architectures and training sets. Moreover, this view yields a simple and fast method of generating adversarial examples. Using this approach to provide examples for adversarial training, we reduce the test set error of a maxout network on the MNIST dataset.

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="1412.6572")

Code

Syntology Ran 10 of 21 code samples harvested from 8 repositories linked to this paper; 11 have no recorded run. Of those that ran: 2 ran · honoured contract; 1 ran · violated contract; 6 ran · our draft was wrong; 1 ran · fixture could not drive it.

By repository: community (archive-listed): 21 samples from 8 repositories, 10 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

59 repositories listed; official and paper-mentioned ones first.

1Konny/FGSM mentioned on GitHubpytorch report
Anaststam/Adversarial-Attacks mentioned on GitHubpytorch report
AngusG/cleverhans-attacking-bnns mentioned on GitHubtfMIT report
BendeguzToth/Fun-with-ConvNets mentioned on GitHubMIT report
HowToMakeABomb101/Hot2MakeAB0mbSite mentioned on GitHubtfMIT report
Jupetus/ExplainableAI mentioned on GitHubpytorch report
KaidongLi/pytorch-LatticePointClassifier mentioned on GitHubpytorchMIT report
LamaLenny/Adversarial-Attack mentioned on GitHub report
LawrenceMMStewart/Adversarial_Attack mentioned on GitHubpytorch report
OwenSec/DeepDetector mentioned on GitHubtf report
SifatMd/Research-Papers mentioned on GitHub report
Trustworthy-AI-Group/TransferAttack mentioned on GitHubpytorch report
amerch/CIFAR100-Training mentioned on GitHubpytorch report
anirudh9784/Adversarial-Defense mentioned on GitHubtf report
anirudh9784/Major_Project mentioned on GitHubtf report
arobey1/advbench mentioned on GitHubpytorch report
bingcheng45/hnr-extension mentioned on GitHubtf report
cfinlay/tulip mentioned on GitHubpytorchMIT report
ckerce/pops_ml mentioned on GitHubpytorch report
cleverhans-lab/cleverhans mentioned on GitHubtfMIT report
coallaoh/whitenblackbox mentioned on GitHubpytorchMIT report
drewbarot/Un-CNN mentioned on GitHubtf report
eiriniOG/seedtag-codetest mentioned on GitHubtf report
elijahcn/TextCNN-AdversarialTraining mentioned on GitHubpytorchMIT report
elites2k19/prism-attack mentioned on GitHubtfMIT report
eth-sri/diffai mentioned on GitHubpytorch report
facebookresearch/adversarial_image_defenses mentioned on GitHubpytorchNOASSERTION report
formal-verification-research/NJSMA mentioned on GitHubtfMIT report
formal-verification-research/verapak mentioned on GitHubtfMIT report
gauthiercler/adversarial-mnist mentioned on GitHubpytorch report
henry8527/GCE mentioned on GitHubpytorch report
iirishikaii/cleverhans mentioned on GitHubtfMIT report
inhopark94/ihpark mentioned on GitHubpytorch report
jaypmorgan/Adversarial.jl mentioned on GitHubpytorch report
jrguo/FastGradientSignMNIST mentioned on GitHubtf report
katiashh/ioi-attack mentioned on GitHubpytorch report
locuslab/convex_adversarial mentioned on GitHubpytorch report
mkazmier/pytorch-fgsm-simple mentioned on GitHubpytorch report
openai/cleverhans mentioned on GitHubtf report
sdemyanov/ConvNet mentioned on GitHubtf report
shijiel2/cleverhans mentioned on GitHubtf report
soumyac1999/FGSM-Keras mentioned on GitHubtf report
tensorflow/cleverhans mentioned on GitHubtfMIT report
winycg/HCGNet mentioned on GitHubpytorch report
yaq007/cleverhans mentioned on GitHubtfMIT report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

21 samples harvested; 10 ran; 2 honoured the contract we drafted; 11 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

2ran · honoured contract
1ran · violated contract
6ran · our draft was wrong
1ran · fixture could not drive it
11unverified

Licence: 8 of the 21 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 8 repositories linked to this paper, official or community; each sample names its own and says which. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

arg_type shijiel2/cleverhans/cleverhans/attacks/attack.py community (archive-listed) ran · our draft was wrong MIT (permissive) · 50b95028f0c76d76 · report
catNonNullErrors eth-sri/diffai/ai.py community (archive-listed) ran · our draft was wrong MIT (permissive) · f8f37cd8a52870c1 · report
conv3x3 amerch/CIFAR100-Training/models/resnet.py community (archive-listed) ran · our draft was wrong no licence file found · pointer only · 583f9780bdd00a45 · report
fast_gradient_attack Anaststam/Adversarial-Attacks/attacks.py community (archive-listed) ran · fixture could not drive it no licence file found · pointer only · 10b255fd33ae2074 · report
fgsm LawrenceMMStewart/Adversarial_Attack/FGSM_LeNet5.py community (archive-listed) ran · our draft was wrong fingerprinted no licence file found · pointer only · 4cb25a22c32d9e0a · report
get_gan_labels KeAWang/BayesianGAN4AdversarialAttacks/bayesian_gan_hmc.py community (archive-listed) ran · violated contract fingerprinted licence not identified · pointer only · a0cc7a40deacf582 · report
get_supervised_batches KeAWang/BayesianGAN4AdversarialAttacks/bayesian_gan_hmc.py community (archive-listed) ran · our draft was wrong licence not identified · pointer only · c3eb6809b747e0a0 · report
get_test_batches KeAWang/BayesianGAN4AdversarialAttacks/bayesian_gan_hmc.py community (archive-listed) ran · our draft was wrong licence not identified · pointer only · 428f94db5821f119 · report
predict_model Anaststam/Adversarial-Attacks/training_and_evaluation.py community (archive-listed) ran · honoured contract no licence file found · pointer only · 81a5ee879107fc8b · report
train_model Anaststam/Adversarial-Attacks/training_and_evaluation.py community (archive-listed) ran · honoured contract no licence file found · pointer only · 9b17713b7849e4b3 · report
creluBoxy eth-sri/diffai/ai.py community (archive-listed) unverified MIT (permissive) · a49c0dd16152c346 · report
creluBoxySound eth-sri/diffai/ai.py community (archive-listed) unverified MIT (permissive) · c42ef7b9012126aa · report
get_onnx_node_shape formal-verification-research/verapak/verapak/model_tools/onnx_utils.py community (archive-listed) unverified MIT (permissive) · ab24078cd09fe35d · report
indices BendeguzToth/Fun-with-ConvNets/project/layers.py community (archive-listed) unverified MIT (permissive) · 3f8a2efde87b7214 · report
load_cifar10 BendeguzToth/Fun-with-ConvNets/project/utils.py community (archive-listed) unverified MIT (permissive) · 4bbe5492f940117b · report
load_graph_by_type formal-verification-research/verapak/verapak/model_tools/model_base.py community (archive-listed) unverified MIT (permissive) · 6ffa680aced49b5f · report
one_hot BendeguzToth/Fun-with-ConvNets/project/utils.py community (archive-listed) unverified MIT (permissive) · df76aed2eb8e5d62 · report
output_to_loss formal-verification-research/verapak/verapak/model_tools/model_base.py community (archive-listed) unverified MIT (permissive) · a7d275667792ebe3 · report
point_to_domain formal-verification-research/verapak/verapak/snap.py community (archive-listed) unverified MIT (permissive) · 0246559ef6f30ca2 · report
region_to_domain formal-verification-research/verapak/verapak/snap.py community (archive-listed) unverified MIT (permissive) · 6a50db6778426abc · report
unpickle BendeguzToth/Fun-with-ConvNets/project/utils.py community (archive-listed) unverified MIT (permissive) · 46842ba14f312456 · report

Tasks

Image Classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification MNIST Explaining and Harnessing Adversarial Examples Percentage error 0.8 #46 of 81 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections