# Towards Data-Free Model Stealing in a Hard Label Setting

Source: [Towards Data-Free Model Stealing in a Hard Label Setting](<https://www.sota2.com/research/paper/68fa6c621e209131cea33bd9>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-04T05:17:11.970Z

Authors: Sunandini Sanyal, Sravanti Addepalli, R. Venkatesh Babu

Published: 2022-04-23T08:44:51.000Z

## Abstract

Machine learning models deployed as a service (MLaaS) are susceptible to model stealing attacks, where an adversary attempts to steal the model within a restricted access framework. While existing attacks demonstrate near-perfect clone-model performance using softmax predictions of the classification network, most of the APIs allow access to only the top-1 labels. In this work, we show that it is indeed possible to steal Machine Learning models by accessing only top-1 predictions (Hard Label setting) as well, without access to model gradients (Black-Box setting) or even the training dataset (Data-Free setting) within a low query budget. We propose a novel GAN-based framework that trains the student and generator in tandem to steal the model effectively while overcoming the challenge of the hard label setting by utilizing gradients of the clone network as a proxy to the victim's gradients. We propose to overcome the large query costs associated with a typical Data-Free setting by utilizing publicly available (potentially unrelated) datasets as a weak image prior. We additionally show that even in the absence of such data, it is possible to achieve state-of-the-art results within a low query budget using synthetically crafted samples. We are the first to demonstrate the scalability of Model Stealing in a restricted access setting on a 100 class dataset as well.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Model Stealing on CIFAR-10 (test)](<https://www.sota2.com/research/sota/model-stealing-on-cifar-10-test.md>): Model Stealing; CIFAR-10 (test); Synthetic Accuracy (Data-Free): 91.24.

- [Image Classification on CIFAR-100 (Model Cloning/Victim Accuracy)](<https://www.sota2.com/research/sota/image-classification-on-cifar-100-model-cloning-victim-accuracy.md>): Image Classification; CIFAR-100; Victim Accuracy: 78.52.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2204.11022>)

[PDF](<https://arxiv.org/pdf/2204.11022.pdf>)
