Name		Name	Last commit message	Last commit date
Latest commit History 168 Commits
.vscode		.vscode
Figures		Figures
NGT-1.3.3		NGT-1.3.3
cleverhans		cleverhans
dataset		dataset
embed_text		embed_text
man/man1		man/man1
models		models
results		results
tmp		tmp
.gitignore		.gitignore
README.md		README.md
attack_model.py		attack_model.py
basic_models.py		basic_models.py
candidate.py		candidate.py
choose_input.py		choose_input.py
eda.py		eda.py
evaluate.py		evaluate.py
fb_test_y.npy		fb_test_y.npy
filenames.py		filenames.py
handle_text.py		handle_text.py
homoglyphAttack.py		homoglyphAttack.py
lstm.py		lstm.py
mnist_tutorial_cw.py		mnist_tutorial_cw.py
nearestneighbors.py		nearestneighbors.py
phonemeAttack.py		phonemeAttack.py
pipeline.py		pipeline.py
query.py		query.py
rating_list.pkl		rating_list.pkl
requirements.txt		requirements.txt
run.sh		run.sh
sample_submission.csv		sample_submission.csv
select_features.py		select_features.py
test.csv		test.csv
test.py		test.py
testRatings.py		testRatings.py
test_500_075.pkl		test_500_075.pkl
test_attack.py		test_attack.py
test_embed.py		test_embed.py
testytest.py		testytest.py
train.csv		train.csv
train_models.py		train_models.py
visualize.py		visualize.py
words.pkl		words.pkl

Repository files navigation

AML-Toxic

About

Adversarial machine learning project that seeks to improve Google's Perspective API for identifying toxic comments in an effort to reduce its bias and make it more robust to attack.