Dense Associative Memory is Robust to Adversarial Inputs

JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

StudentID *

What did the n value showed in the presented images *

1 point

Number of classes

Number of images in dataset

Power of the interaction vertex

Number of images inside a class

What is the approximate accuracy of DAM models on MNIST dataset? *

1 point

99.5%

98.5%

95%

90%

Which n values were used for training the models *

1 point

n=1

n=10

n=20

n=30

Required

Models with high n value are better than models with small n value in *

1 point

Mimicking human visual cognition

Extracting features from images

Getting considerably better accuracy on MNIST dataset

Being less confident when predicting adversarial images

Required

Submit

Clear form

Never submit passwords through Google Forms.

This content is neither created nor endorsed by Google. - Terms of Service - Privacy Policy

Does this form look suspicious? Report