Increasing the cost of model extraction with calibrated proof of work

A Dziedzic, MA Kaleem, YS Lu, N Papernot - arXiv preprint arXiv …, 2022 - arxiv.org
In model extraction attacks, adversaries can steal a machine learning model exposed via a
public API by repeatedly querying it and adjusting their own model based on obtained …
被引用次数：27 相关文章
[PDF] researchgate.net

[PDF][PDF] INCREASING THE COST OF MODEL EXTRACTION WITH CALIBRATED PROOF OF WORK

A Dziedzic, MA Kaleem, YS Lu, N Papernot - researchgate.net
In model extraction attacks, adversaries can steal a machine learning model exposed via a
public API by repeatedly querying it and adjusting their own model based on obtained …
[PDF] adam-dziedzic.com

[PDF][PDF] INCREASING THE COST OF MODEL EXTRACTION WITH CALIBRATED PROOF OF WORK

A Dziedzic, MA Kaleem, YS Lu, N Papernot - adam-dziedzic.com
In model extraction attacks, adversaries can steal a machine learning model exposed via a
public API by repeatedly querying it and adjusting their own model based on obtained …

Increasing the Cost of Model Extraction with Calibrated Proof of Work

A Dziedzic, MA Kaleem, YS Lu, N Papernot - International Conference on … - openreview.net
In model extraction attacks, adversaries can steal a machine learning model exposed via a
public API by repeatedly querying it and adjusting their own model based on obtained …

Increasing the Cost of Model Extraction with Calibrated Proof of Work

A Dziedzic, MA Kaleem, Y Shen Lu… - arXiv e …, 2022 - ui.adsabs.harvard.edu
In model extraction attacks, adversaries can steal a machine learning model exposed via a
public API by repeatedly querying it and adjusting their own model based on obtained …