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

Recent work has found that adversarially-robust deep networks used for imageclassification are more interpretable: their feature attributions tend to besharper, and are more concentrated on the objects associated with the image'sground-truth class. We show that smooth decision boundaries play an importantrole in this enhanced interpretability, as the model's input gradients arounddata points will more closely align with boundaries' normal vectors when theyare smooth. Thus, because robust models have smoother boundaries, the resultsof gradient-based attribution methods, like Integrated Gradients and DeepLift,will capture more accurate information about nearby decision boundaries. Thisunderstanding of robust interpretability leads to our second contribution:\emph{boundary attributions}, which aggregate information about the normalvectors of local decision boundaries to explain a classification outcome. Weshow that by leveraging the key factors underpinning robust interpretability,boundary attributions produce sharper, more concentrated visual explanations --even on non-robust models. Any example implementation can be found at\url{https://github.com/zifanw/boundary}.

Robust Models Are More Interpretable Because Attributions Look Normal