SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.10162 · ICLR · 2020

EMPIR: Ensembles of Mixed Precision Deep Networks for Increased Robustness against Adversarial Attacks

Sanchari Sen, Balaraman Ravindran, Anand Raghunathan

arXiv · PDF · Open in the Atlas

Code that ran

5 of 17 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 0 failures of our instrument; 12 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2004.10162.json · badge: /api/ran/2004.10162.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
sancharisen/EMPIR canonical (classifier label)
classifier basis: census
5 ran · 12 no recorded run
FunctionStatusWhere it lives
batch_indices Ran
ran, honoured the contract we drafted
sancharisen/EMPIR/cleverhans/utils.py
pointer only (licence: MIT) · get_code("96b581b6ac010981")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

hard_sigmoid Ran
ran on a synthesized input
sancharisen/EMPIR/cleverhans_tutorials/tutorial_models.py
pointer only (licence: MIT) · get_code("33ce7291d7174c33")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

model_loss Ran
ran on a synthesized input
sancharisen/EMPIR/cleverhans/utils_tf.py
pointer only (licence: MIT) · get_code("24e8cacf2e978687")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

other_classes Ran
ran, honoured the contract we drafted
sancharisen/EMPIR/cleverhans/utils.py
pointer only (licence: MIT) · get_code("563a92a1c7232061")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

to_categorical Ran
ran on a synthesized input
sancharisen/EMPIR/cleverhans/utils.py
pointer only (licence: MIT) · get_code("436b1ff1875d0cc7")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

cnn_model Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/utils_keras.py
pointer only (licence: MIT) · get_code("69dbec087f603dc2")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

conv_2d Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/utils_keras.py
pointer only (licence: MIT) · get_code("4e8d1df070cb5f83")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

data_mnist Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/utils_mnist.py
pointer only (licence: MIT) · get_code("5f6a796fa47c9ff8")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

fgm Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/attacks_tf.py
pointer only (licence: MIT) · get_code("46c4fbb3b28111b4")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

fgsm Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/attacks.py
pointer only (licence: MIT) · get_code("669a4a70d107c4a4")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

fgsm Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/attacks_tf.py
pointer only (licence: MIT) · get_code("ab300465511282c2")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

jsma Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/attacks.py
pointer only (licence: MIT) · get_code("48e3be63336a0ac5")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

normalized_sign_changes_op Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/utils_tf.py
pointer only (licence: MIT) · get_code("457d5efa3991f205")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

quantize_grad Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans_tutorials/tutorial_models.py
pointer only (licence: MIT) · get_code("88e4e6b729fd7669")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

sign_changes_count_op Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/utils_tf.py
pointer only (licence: MIT) · get_code("45a3aac96deaa69d")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

vatm Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/attacks.py
pointer only (licence: MIT) · get_code("265a717d9cb4752d")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

vatm Not yet run
unverified (harvested, no recorded run)
sancharisen/EMPIR/cleverhans/attacks_tf.py
pointer only (licence: MIT) · get_code("1d76eb5891f63a2f")
How it ran

The body of this sample is withheld: MIT, as GitHub detects it on this paper's repository, sancharisen/EMPIR, is in the permissive set, but a licence decision keeps the body a pointer: the reason for that decision is not recorded with it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 17 distinct functions with an identity, 17 are also present in other papers whose code we hold (17 with at least one paper dated earlier, 0 the same month, 2 later) and 0 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
batch_indices 4 other papers
2 dated earlier, 2 later
1610.00768 (2016-10)
Technical Report on the CleverHans v2.1.0 Adversarial Examples Library
1711.00449 (2017-11)
Attacking Binarized Neural Networks
2006.16375 (2020-06)
Improving Calibration through the Relationship with Adversarial Robustness
and 1 more not named here
other_classes 4 other papers
2 dated earlier, 2 later
1610.00768 (2016-10)
Technical Report on the CleverHans v2.1.0 Adversarial Examples Library
1711.00449 (2017-11)
Attacking Binarized Neural Networks
2006.16375 (2020-06)
Improving Calibration through the Relationship with Adversarial Robustness
and 1 more not named here
cnn_model 2 other papers
2 dated earlier
1610.00768 (2016-10)
Technical Report on the CleverHans v2.1.0 Adversarial Examples Library
1711.00449 (2017-11)
Attacking Binarized Neural Networks
conv_2d 2 other papers
2 dated earlier
1610.00768 (2016-10)
Technical Report on the CleverHans v2.1.0 Adversarial Examples Library
1711.00449 (2017-11)
Attacking Binarized Neural Networks
data_mnist 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
fgm 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
fgsm 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
fgsm 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
hard_sigmoid 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
jsma 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
model_loss 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
normalized_sign_changes_op 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
quantize_grad 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
sign_changes_count_op 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
to_categorical 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
vatm 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks
vatm 1 other paper
1 dated earlier
1711.00449 (2017-11)
Attacking Binarized Neural Networks

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

Ensuring robustness of Deep Neural Networks (DNNs) is crucial to their adoption in safety-critical applications such as self-driving cars, drones, and healthcare. Notably, DNNs are vulnerable to adversarial attacks in which small input perturbations can produce catastrophic misclassifications. In this work, we propose EMPIR, ensembles of quantized DNN models with different numerical precisions, as a new approach to increase robustness against adversarial attacks. EMPIR is based on the observation that quantized neural networks often demonstrate much higher robustness to adversarial attacks than full precision networks, but at the cost of a substantial loss in accuracy on the original (unperturbed) inputs. EM-PIR overcomes this limitation to achieve the "best of both worlds", i.e., the higher unperturbed accuracies of the full precision models combined with the higher robustness of the low precision models, by composing them in an ensemble. Further, as low precision DNN models have significantly lower computational and storage requirements than full precision models, EMPIR models only incur modest compute and memory overheads compared to a single full-precision model (<25% in our evaluations). We evaluate EMPIR across a suite of DNNs for 3 different image recognition tasks (MNIST, CIFAR-10 and ImageNet) and under 4 different adversarial attacks. Our results indicate that EMPIR boosts the average adversarial accuracies by 42.6%, 15.2% and 10.5% for the DNN models trained on the MNIST, CIFAR-10 and ImageNet datasets respectively, when compared to single full-precision models, without sacrificing accuracy on the unperturbed inputs.

For agents

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

get_harvested_code_for_paper("2004.10162")
get_code_for_paper("2004.10162")
have("2004.10162")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2004.10162.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2004.10162.svg)](https://syntology.ai/paper/2004.10162)

Connect an agent — have() is free.