English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

For a standard convolutional neural network, optimizing over the input pixelsto maximize the score of some target class will generally produce agrainy-looking version of the original image. However, Santurkar et al. (2019)demonstrated that for adversarially-trained neural networks, this optimizationproduces images that uncannily resemble the target class. In this paper, weshow that these "perceptually-aligned gradients" also occur under randomizedsmoothing, an alternative means of constructing adversarially-robustclassifiers. Our finding supports the hypothesis that perceptually-alignedgradients may be a general property of robust classifiers. We hope that ourresults will inspire research aimed at explaining this link betweenperceptually-aligned gradients and adversarial robustness.

Are Perceptually-Aligned Gradients a General Property of Robust Classifiers?