SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1906.01444 · 2019

Heterogeneous Gaussian Mechanism: Preserving Differential Privacy in Deep Learning with Provable Robustness

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 15 functions out of this paper's own repositories and ran 5 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
haiphanNJIT/SecureSGD canonical 3 of 3
haiphanNJIT/StoBatch pwc_unofficial 2 of 12
FunctionStatusWhere it lives
avg_pool Ran haiphanNJIT/StoBatch/StoBatchCIFAR10/StoBatch_cifar10.py
code served (permissive licence) · get_code("67a83a4977e5900b")
comp_sigma Ran haiphanNJIT/SecureSGD/MNIST/SecureSGD.py
pointer only (licence: NONE) · get_code("5a277c84f6cc91ea")
compute_sigma Ran haiphanNJIT/SecureSGD/MNIST/SecureSGD.py
pointer only (licence: NONE) · get_code("c8a866778827a03d")
fixed_padding Ran haiphanNJIT/StoBatch/StoBatchTinyImageNet/resnet_utils.py
code served (permissive licence) · get_code("7fa95a4ad6afe798")
parse_time Ran haiphanNJIT/SecureSGD/MNIST/SecureSGD.py
pointer only (licence: NONE) · get_code("0d51b5ad039c5e1b")
batch_norm Not yet run haiphanNJIT/StoBatch/StoBatchTinyImageNet/resnet_utils.py
code served (permissive licence) · get_code("29e0dcee4d607b62")
bias_variable Not yet run haiphanNJIT/StoBatch/MNIST/StoBatch.py
code served (permissive licence) · get_code("aafb9bf91f6ef802")
conv2d Not yet run haiphanNJIT/StoBatch/CIFAR10/StoBatch.py
code served (permissive licence) · get_code("2c2112993af8d032")
conv2d_fixed_padding Not yet run haiphanNJIT/StoBatch/StoBatchTinyImageNet/resnet_utils.py
code served (permissive licence) · get_code("60ed1730c5f4efa0")
generateIdLMNoise Not yet run haiphanNJIT/StoBatch/StoBatchTinyImageNet/StoBatch_resnet_pretrain.py
code served (permissive licence) · get_code("991a7403637402a8")
generateNoise Not yet run haiphanNJIT/StoBatch/StoBatchTinyImageNet/StoBatch_resnet_pretrain.py
code served (permissive licence) · get_code("9c54b765a044a6b6")
lossDPSGD Not yet run haiphanNJIT/StoBatch/StoBatchTinyImageNet/SSGD_loss.py
code served (permissive licence) · get_code("6764ccff0ce428ed")
max_out Not yet run haiphanNJIT/StoBatch/CIFAR10/StoBatch.py
code served (permissive licence) · get_code("b7a8162db3a2edfb")
parametric_relu Not yet run haiphanNJIT/StoBatch/CIFAR10/StoBatch.py
code served (permissive licence) · get_code("9a523e5bd9c79de4")
weight_variable Not yet run haiphanNJIT/StoBatch/MNIST/StoBatch.py
code served (permissive licence) · get_code("2a775c69ec858b1e")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

In this paper, we propose a novel Heterogeneous Gaussian Mechanism (HGM) to preserve differential privacy in deep neural networks, with provable robustness against adversarial examples. We first relax the constraint of the privacy budget in the traditional Gaussian Mechanism from (0, 1] to (0, \infty), with a new bound of the noise scale to preserve differential privacy. The noise in our mechanism can be arbitrarily redistributed, offering a distinctive ability to address the trade-off between model utility and privacy loss. To derive provable robustness, our HGM is applied to inject Gaussian noise into the first hidden layer. Then, a tighter robustness bound is proposed. Theoretical analysis and thorough evaluations show that our mechanism notably improves the robustness of differentially private deep neural networks, compared with baseline approaches, under a variety of model attacks.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("1906.01444")
get_code_for_paper("1906.01444")
have("1906.01444")

Connect an agent — have() is free.