SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.14846 · CVPR · 2023

Introducing Competition to Boost the Transferability of Targeted Adversarial Examples through Clean Feature Mixup

Seungju Cho, Changick Kim, Junyoung Byun, Myung-Joon Kwon, Yoonji Kim

arXiv · PDF · Open in the Atlas

Code that ran

8 of 10 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; 0 honoured the contract we drafted, 1 violated it and 3 have no contract verdict; 4 failures of our instrument; 2 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/2305.14846.json · badge: /api/ran/2305.14846.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
dreamflake/cfm canonical (classifier label)
classifier basis: census
8 ran · 1 no recorded run
dreamflake/CFM community
Papers with Code archive
0 ran · 1 no recorded run
FunctionStatusWhere it lives
Linf_PGD Ran
ran; our contract draft was wrong, not the code
dreamflake/cfm/DVERGE/distillation.py
pointer only (licence: MIT); the stored copy get_code serves, from zjysteven/DVERGE, is pointer only (licence: NONE) · get_code("ccdb43ecb62325f2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from zjysteven/DVERGE, and GitHub detected no licence in that repository when it was checked. 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.

compute_rotation Ran
ran on a synthesized input
dreamflake/cfm/attacks.py
code served (permissive licence) · get_code("50956689e2a717ad")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 3, 3) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a batch of angles: shape (B, 3) where B=2
# angles should contain rotation angles in radians for x, y, z axes
angles = torch.tensor([
    [0.1, 0.2, 0.3],
    [0.4, 0.5, 0.6]
], dtype=torch.float32)

ARGS = {
    "angles": angles
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
gradient_wrt_feature Ran
ran; our contract draft was wrong, not the code
dreamflake/cfm/DVERGE/distillation.py
pointer only (licence: MIT); the stored copy get_code serves, from zjysteven/DVERGE, is pointer only (licence: NONE) · get_code("c5ea22297f8dcf56")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from zjysteven/DVERGE, and GitHub detected no licence in that repository when it was checked. 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.

gradient_wrt_input Ran
ran; our contract draft was wrong, not the code
dreamflake/cfm/DVERGE/distillation.py
pointer only (licence: MIT); the stored copy get_code serves, from zjysteven/DVERGE, is pointer only (licence: NONE) · get_code("a7432682e97da06b")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from zjysteven/DVERGE, and GitHub detected no licence in that repository when it was checked. 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.

load_ground_truth Ran
ran; our contract draft was wrong, not the code
dreamflake/cfm/eval_attacks.py
pointer only (licence: MIT); the stored copy get_code serves, from sclbd/transfer_attack_rap, is pointer only (licence: NONE) · get_code("9e4f4d7edbf00ba0")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from sclbd/transfer_attack_rap, and GitHub detected no licence in that repository when it was checked. 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.

load_ground_truth_cifar10 Ran
ran on a synthesized input
dreamflake/cfm/eval_attacks.py
code served (permissive licence) · get_code("ae3730ed44c5ff5b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
import shutil

# Create a temporary directory with mock CIFAR-10 image files
temp_dir = tempfile.mkdtemp()

# Create mock image files with the naming convention expected by the function
# Convention: {image_id}_{original_label}_{target_label}.png
filenames = [
    "img_001_3_7.png",
    "img_002_5_2.png",
    "img_003_0_6.png",
]

for filename in filenames:
    filepath = os.path.join(temp_dir, filename)
    with open(filepath, 'w') as f:
        f.write("")

ARGS = {
    "img_dir": temp_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 2, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rigid_transform Ran
ran on a synthesized input
dreamflake/cfm/attacks.py
code served (permissive licence) · get_code("422be6728d89fc79")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (2, 4, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# vs: vertices tensor, used in matmul(vs, rot)
# Shape: (batch, num_vertices, 3) for 3D points
vs = torch.randn(2, 4, 3)

# rot: rotation matrix, used in matmul(vs, rot)
# Shape: (3, 3) for 3D rotation
rot = torch.randn(3, 3)

# trans: translation vector, used as trans.view(-1, 1, 3)
# Shape: (batch, 3) so it can be viewed as (batch, 1, 3)
trans = torch.randn(2, 3)

ARGS = {
    "vs": vs,
    "rot": rot,
    "trans": trans,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 4, 3) — shape (2, 4, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
str2bool Ran
ran, violated the contract we drafted
dreamflake/cfm/utils.py
code served (permissive licence) · get_code("8605dc8a088f3db8")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned bool, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import argparse

def str2bool(v):
    if isinstance(v, bool):
        return v
    if v.lower() in ('yes', 'true', 't', 'y', '1'):
        return True
    elif v.lower() in ('no', 'false', 'f', 'n', '0'):
        return False
    else:
        raise argparse.ArgumentTypeError('Boolean value expected.')

# Test cases for the str2bool function
ARGS = {
    'v': 'true'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
FeatureMixup Not yet run
unverified (harvested, no recorded run)
dreamflake/CFM/attacks.py
code served (permissive licence) · get_code("8916ca54576c0afd")
How it ran

no invocation recorded.

model_dataset_from_store Not yet run
unverified (harvested, no recorded run)
dreamflake/cfm/MadryLab/model_utils.py
code served (permissive licence) · get_code("c1e41af537b4027c")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 6 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 2 the same month, 2 later) and 4 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
str2bool 175 other papers
68 dated earlier, 7 the same month, 97 later, 3 same year, order not determinable · 2 only as an assembly brick
1406.6247 (2014-06)
Recurrent Models of Visual Attention
1603.09382 (2016-03)
Deep Networks with Stochastic Depth
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
and 172 more not named here
model_dataset_from_store 4 other papers
1 dated earlier, 1 the same month, 2 later
2106.02078 (2021-06)
Improving Neural Network Robustness via Persistency of Excitation
2305.19101 (2023-05)
Which Models have Perceptually-Aligned Gradients? An Explanation via Off-Manifold Robustness
2405.08886 (2024-05)
The Pitfalls and Promise of Conformal Inference Under Adversarial Attacks
and 1 more not named here
load_ground_truth 3 other papers
3 dated earlier
2012.11207 (2020-12)
On Success and Simplicity: A Second Look at Transferable Targeted Attacks
2103.14641 (2021-03)
On Generating Transferable Targeted Perturbations
2210.05968 (2022-10)
Boosting the Transferability of Adversarial Attacks with Reverse Adversarial Perturbation
Linf_PGD 1 other paper
1 dated earlier
2009.14720 (2020-09)
DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensembles
gradient_wrt_feature 1 other paper
1 dated earlier
2009.14720 (2020-09)
DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensembles
gradient_wrt_input 1 other paper
1 dated earlier
2009.14720 (2020-09)
DVERGE: Diversifying Vulnerabilities for Enhanced Robust Generation of Ensembles

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

Deep neural networks are widely known to be susceptible to adversarial examples, which can cause incorrect predictions through subtle input modifications. These adversarial examples tend to be transferable between models, but targeted attacks still have lower attack success rates due to significant variations in decision boundaries. To enhance the transferability of targeted adversarial examples, we propose introducing competition into the optimization process. Our idea is to craft adversarial perturbations in the presence of two new types of competitor noises: adversarial perturbations towards different target classes and friendly perturbations towards the correct class. With these competitors, even if an adversarial example deceives a network to extract specific features leading to the target class, this disturbance can be suppressed by other competitors. Therefore, within this competition, adversarial examples should take different attack strategies by leveraging more diverse features to overwhelm their interference, leading to improving their transferability to different models. Considering the computational complexity, we efficiently simulate various interference from these two types of competitors in feature space by randomly mixing up stored clean features in the model inference and named this method Clean Feature Mixup (CFM). Our extensive experimental results on the ImageNet-Compatible and CIFAR-10 datasets show that the proposed method outperforms the existing baselines with a clear margin. Our code is available at https://github.com/dreamflake/CFM .

For agents

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

get_harvested_code_for_paper("2305.14846")
get_code_for_paper("2305.14846")
have("2305.14846")

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

curl https://syntology.ai/api/ran/2305.14846.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.14846.svg)](https://syntology.ai/paper/2305.14846)

Connect an agent — have() is free.