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

Model extraction attacks have renewed interest in the classic problem oflearning neural networks from queries. In this work we give the firstpolynomial-time algorithm for learning arbitrary one hidden layer neuralnetworks activations provided black-box access to the network. Formally, weshow that if $F$ is an arbitrary one hidden layer neural network with ReLUactivations, there is an algorithm with query complexity and running time thatis polynomial in all parameters that outputs a network $F'$ achieving lowsquare loss relative to $F$ with respect to the Gaussian measure. While anumber of works in the security literature have proposed and empiricallydemonstrated the effectiveness of certain algorithms for this problem, ours isthe first with fully polynomial-time guarantees of efficiency even forworst-case networks (in particular our algorithm succeeds in theoverparameterized setting).

Efficiently Learning Any One Hidden Layer ReLU Network From Queries