search

nesl / nlp_adversarial_examples

Implementation code for the paper "Generating Natural Language Adversarial Examples"
MIT License
167 stars 49 forks source link

Confused about baseline attack #6

Open xiaotili0123 opened 4 years ago

xiaotili0123 commented 4 years ago

Hi there,

I am highly interested in your fantastic work. However, I am a little bit confused about the baseline attack you used in your code and your paper. In BaselineDemo.ipynb, what is the NewBaseline attack that you imported? I couldn't find it in Attacks.py but PerturbSentBaseline(). Also, you mentioned in your paper that you use a baseline method to greedily find adversarial texts, could you please be more specific on the baseline attack you use? Thank you so much.

uestc-hjw commented 4 years ago

hello, i want to know,run the "python compute_dist_mat.py",why the programes could show the words"have killed"?

jxmorris12 commented 4 years ago

Hi @Crystallee612 and @falllenleaves -- you might be interested in trying your experiments in TextAttack instead. TextAttack's 'alzantot' recipe is a reimplementation of this paper with some performance enhancements that's been endorsed by Alzantot (the creator of this work). TextAttack doesn't implement the baseline, but we'd be happy to work with you to add it. Hope that helps!