Polynomial Time Cryptanalytic Extraction of Neural Network Models

Training a 3-node neural network is np-complete

Avrim Blum and Ronald L. Rivest · 1993

Earlier work this paper cites.

Reconstructing a neural net from its output

Charles Fefferman · 1994

Earlier work this paper cites.

Adversarial learning

Daniel Lowd and Christopher Meek · 2005

Earlier work this paper cites.

80 million tiny images: A large data set for nonparametric object and scene recognition

Antonio Torralba, Rob Fergus, and William T Freeman · 2008

Earlier work this paper cites.

How far can we go without convolution: Improving fully-connected networks

Original

Zhouhan Lin, Roland Memisevic, and Kishore Konda · 2015

Earlier work this paper cites.

Stealing machine learning models via prediction { \{ APIs } \}

Original

Florian Tramèr, Fan Zhang, Ari Juels, Michael K Reiter, and Thomas Ristenpart · 2016

Earlier work this paper cites.

Practical black-box attacks against machine learning

Nicolas Papernot, Patrick D. McDaniel, Ian J. Goodfellow, Somesh Jha, Z. Berkay Celik, and Ananthram Swami · 2017

Cited alongside, same era.

CSI NN: reverse engineering of neural network architectures through electromagnetic side channel

Lejla Batina, Shivam Bhasin, Dirmanto Jap, and Stjepan Picek · 2019

Cited alongside, same era.

Model reconstruction from model explanations

Smitha Milli, Ludwig Schmidt, Anca D. Dragan, and Moritz Hardt · 2019

Cited alongside, same era.

Efficiently stealing your machine learning models

Robert Nikolai Reith, Thomas Schneider, and Oleksandr Tkachenko · 2019

Cited alongside, same era.

Cryptanalytic extraction of neural network models

Nicholas Carlini, Matthew Jagielski, and Ilya Mironov · 2020

Cited alongside, same era.

An image is worth 16x16 words: Transformers for image recognition at scale

High accuracy and high fidelity extraction of neural networks

Matthew Jagielski, Nicholas Carlini, David Berthelot, Alex Kurakin, and Nicolas Papernot · 2020

Later among the works it cites.

Reverse-engineering deep relu networks

David Rolnick and Konrad P. Körding · 2020

Later among the works it cites.

An exact poly-time membership-queries algorithm for extraction a three-layer relu network

Original

Amit Daniely and Elad Granot · 2021

Later among the works it cites.

A practical introduction to side-channel extraction of deep neural network parameters

Raphaël Joud, Pierre-Alain Moëllic, Simon Pontié, and Jean-Baptiste Rigaud · 2022

Later among the works it cites.

I know what you trained last summer: A survey on stealing machine learning models and defences

Daryna Oliynyk, Rudolf Mayer, and Andreas Rauber · 2023

Closest in time.