Papers › Why Does Little Robustness Help? A Further Step Towards Understanding Adversarial...

Why Does Little Robustness Help? A Further Step Towards Understanding Adversarial Transferability

15 Jul 2023arXiv:2307.07873archive 2025-07-28

Yechao Zhang, Shengshan Hu, Leo Yu Zhang, Junyu Shi, Minghui Li, Xiaogeng Liu, Wei Wan, Hai Jin

Adversarial examples (AEs) for DNNs have been shown to be transferable: AEs that successfully fool white-box surrogate models can also deceive other black-box models with different architectures. Although a bunch of empirical studies have provided guidance on generating highly transferable AEs, many of these findings lack explanations and even lead to inconsistent advice. In this paper, we take a further step towards understanding adversarial transferability, with a particular focus on surrogate aspects. Starting from the intriguing little robustness phenomenon, where models adversarially trained with mildly perturbed adversarial samples can serve as better surrogates, we attribute it to a trade-off between two predominant factors: model smoothness and gradient similarity. Our investigations focus on their joint effects, rather than their separate correlations with transferability. Through a series of theoretical and empirical analyses, we conjecture that the data distribution shift in adversarial training explains the degradation of gradient similarity. Building on these insights, we explore the impacts of data augmentation and gradient regularization on transferability and identify that the trade-off generally exists in the various training mechanisms, thus building a comprehensive blueprint for the regulation mechanism behind transferability. Finally, we provide a general route for constructing better surrogates to boost transferability which optimizes both model smoothness and gradient similarity simultaneously, e.g., the combination of input gradient regularization and sharpness-aware minimization (SAM), validated by extensive experiments. In summary, we call for attention to the united impacts of these two factors for launching effective transfer attacks, rather than optimizing one while ignoring the other, and emphasize the crucial role of manipulating surrogate models.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 15 code samples harvested from 1 repository linked to this paper: 12 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 6 where Syntology's instrument failed) · 3 unverified. Where Syntology's instrument failed: 6 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 15 samples from 1 repository, 12 ran (6 with no instrument failure; 6 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 12 samples that ran has no recorded date.

cgcl-codes/transferattacksurrogates officialmentioned in paperpytorchMITSyntology: 15 samples, 12 ran (6 instrument-failed) · 3 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

15 samples harvested: 12 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 0 honoured, 0 violated, 6 with no contract checked; 6 where Syntology's instrument failed) · 3 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

6ran · our draft was wrong
6ran
3unverified

Licence: 7 of the 15 samples are pointer only, meaning Syntology does not serve that copy's text. For 7 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from cgcl-codes/transferattacksurrogates. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

cgcl-codes/transferattacksurrogates official repository 15 samples, 12 ran (6 instrument-failed) · 3 unverified
clip_by_tensor in 6 papers TransferAttack/imagenet_attack.py ran · our draft was wrong fingerprinted MIT (permissive) · 11d8e1b97b2f5801 · report fixture: signature t:Tensor/2/float, t_min:Tensor/0/float, t_max:Tensor/0/float · returned Tensor · level recorded, date not recorded
drop_connect in 73 papers
census: 9 implementations from 7 papers, 4 distinct outputs across 2 buckets, one shared input per bucket
TransferAttack/models/efficientnet.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 4304a326c593f8db · report fixture: signature x:Tensor/4/float, drop_ratio:float/0/float · returned Tensor · level recorded 2026-09-24
make_layers in 155 papers TransferAttack/models/vgg.py ran MIT (permissive) here; the stored copy is pointer only · 424b4090e4679330 · report fixture: signature cfg:list/1/int, in_dims:int/0/int, batch_norm:bool/0/bool · returned Sequential · level recorded 2026-09-24
resnet18 in 432 papers TransferAttack/models/resnet.py ran · our draft was wrong MIT (permissive) · 33b1291f4ed17ae8 · report fixture: signature in_dims:int/0/int, out_dims:int/0/int · returned ResNet · level recorded 2026-09-24
resnet20 in 42 papers TransferAttack/models/resnet_cifar.py ran MIT (permissive) · ef2cb901ccb65deb · report fixture: signature num_classes:int/0/int · returned ResNet · level recorded 2026-09-24
resnet32 in 47 papers TransferAttack/models/resnet_cifar.py ran MIT (permissive) · 580a194529286f83 · report fixture: signature num_classes:int/0/int · returned ResNet · level recorded 2026-09-24
resnet44 in 19 papers TransferAttack/models/resnet_cifar.py ran MIT (permissive) · 7a0631b66b62efa7 · report fixture: signature num_classes:int/0/int · returned ResNet · level recorded 2026-09-24
resnet50 in 121 papers TransferAttack/models/resnet.py ran · our draft was wrong MIT (permissive) · 8a672b592867c614 · report fixture: signature in_dims:int/0/int, out_dims:int/0/int · returned ResNet · level recorded 2026-09-24
swish in 258 papers
census: 14 implementations from 37 papers, 2 distinct outputs on one shared input
TransferAttack/models/efficientnet.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 8737c82de631cffc · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-24
vgg11_bn in 40 papers TransferAttack/models/vgg.py ran MIT (permissive) here; the stored copy is pointer only · acc8c58f222a7df0 · report fixture: signature in_dims:int/0/int, out_dims:int/0/int · returned VGG · level recorded 2026-09-24
vgg13 in 10 papers TransferAttack/models/vgg.py ran MIT (permissive) · 573861d5c02def4d · report fixture: signature out_channels:int/0/int · returned VGG · level recorded 2026-09-24
wrn34_10 TransferAttack/models/resnet.py ran · our draft was wrong MIT (permissive) · 91926dac2d2c5373 · report fixture: signature in_dims:int/0/int, out_dims:int/0/int · returned WRN · level recorded 2026-09-24
densenet121 in 43 papers TransferAttack/models/densenet.py unverified MIT (permissive) here; the stored copy is pointer only · 7510d0bcd890c346 · report
densenet161 in 20 papers TransferAttack/models/densenet.py unverified MIT (permissive) here; the stored copy is pointer only · 8fc5580681158f30 · report
densenet169 in 42 papers TransferAttack/models/densenet.py unverified MIT (permissive) here; the stored copy is pointer only · 54ca014637609170 · report

Tasks

AttributeData Augmentation

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

FocusSharpness-Aware Minimization

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections