Papers › Robust Weight Signatures: Gaining Robustness as Easy as Patching Weights?

Robust Weight Signatures: Gaining Robustness as Easy as Patching Weights?

24 Feb 2023arXiv:2302.12480archive 2025-07-28

Ruisi Cai, Zhenyu Zhang, Zhangyang Wang

Given a robust model trained to be resilient to one or multiple types of distribution shifts (e.g., natural image corruptions), how is that "robustness" encoded in the model weights, and how easily can it be disentangled and/or "zero-shot" transferred to some other models? This paper empirically suggests a surprisingly simple answer: linearly - by straightforward model weight arithmetic! We start by drawing several key observations: (1)assuming that we train the same model architecture on both a clean dataset and its corrupted version, resultant weights mostly differ in shallow layers; (2)the weight difference after projection, which we call "Robust Weight Signature" (RWS), appears to be discriminative and indicative of different corruption types; (3)for the same corruption type, the RWSs obtained by one model architecture are highly consistent and transferable across different datasets. We propose a minimalistic model robustness "patching" framework that carries a model trained on clean data together with its pre-extracted RWSs. In this way, injecting certain robustness to the model is reduced to directly adding the corresponding RWS to its weight. We verify our proposed framework to be remarkably (1)lightweight. since RWSs concentrate on the shallowest few layers and we further show they can be painlessly quantized, storing an RWS is up to 13 x more compact than storing the full weight copy; (2)in-situ adjustable. RWSs can be appended as needed and later taken off to restore the intact clean model. We further demonstrate one can linearly re-scale the RWS to control the patched robustness strength; (3)composable. Multiple RWSs can be added simultaneously to patch more comprehensive robustness at once; and (4)transferable. Even when the clean model backbone is continually adapted or updated, RWSs remain as effective patches due to their outstanding cross-dataset transferability.

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 19 code samples harvested from 1 repository linked to this paper: 11 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 3 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 19 samples from 1 repository, 11 ran (8 with no instrument failure; 3 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 between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 2 of the 11 samples that ran have no recorded date.

VITA-Group/Robust_Weight_Signatures officialmentioned on GitHubpytorchMITSyntology: 19 samples, 11 ran (3 instrument-failed) · 8 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

19 samples harvested: 11 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 1 violated, 7 with no contract checked; 3 where Syntology's instrument failed) · 8 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · violated contract
3ran · our draft was wrong
7ran
8unverified

Licence: 2 of the 19 samples are pointer only, meaning Syntology does not serve that copy's text. For 2 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 VITA-Group/Robust_Weight_Signatures. “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.

VITA-Group/Robust_Weight_Signatures official repository 19 samples, 11 ran (3 instrument-failed) · 8 unverified
affine_quant quantize_utils/quant_utils.py ran MIT (permissive) · 87ed711467d719d3 · report fixture: signature x:Tensor/2/float, scale:Tensor/0/float, zero_point:Tensor/0/float, inplace:bool/0/bool · returned Tensor · level recorded 2026-09-25
attack_pgd in 10 papers trainer/engine_pgd.py ran MIT (permissive) · 28cadfa8650befa7 · report fixture: signature model:callable/0/?, X:Tensor/4/float, y:Tensor/1/int, epsilon:float/0/float, alpha:float/0/float, attack_iters:int/0/int, norm:str/0/str, early_stop:bool/0/bool, restarts:int/0/int · returned Tensor · level recorded 2026-09-25
check_bn in 9 papers tools.py ran · violated contract MIT (permissive) here; the stored copy is pointer only · 6e59364a7cf6008f · report fixture: signature model:callable/0/? · returned bool · level recorded 2026-09-22
conv1x1 in 691 papers models/resnet.py ran · our draft was wrong MIT (permissive) · d9def42110729a85 · report fixture: signature in_planes:int/0/int, out_planes:int/0/int, stride:int/0/int · returned Conv2d · level recorded, date not recorded
conv3x3 in 1,376 papers models/resnet.py ran · our draft was wrong MIT (permissive) · 160bb14bd76201b4 · report fixture: signature in_planes:int/0/int, out_planes:int/0/int, stride:int/0/int, groups:int/0/int, dilation:int/0/int · returned Conv2d · level recorded, date not recorded
get_dirs in 5 papers tools.py ran MIT (permissive) · 3aea45f74b2c92f2 · report fixture: signature root_dir:str/0/str, corruption:list/1/str · returned dict · level recorded 2026-09-25
get_projection in 6 papers tools.py ran fingerprinted MIT (permissive) · 5c47b10f35d34594 · report fixture: signature a:Tensor/1/float, param:Tensor/2/float · returned Tensor · level recorded 2026-09-25
make_layers in 155 papers models/vgg.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · eba3f5bcc6a06d36 · report fixture: signature cfg:list/1/int, batch_norm:bool/0/bool · returned Sequential · level recorded 2026-09-22
signer quantize_utils/quant_utils.py ran fingerprinted MIT (permissive) · 8357390e4d4ad484 · report fixture: signature x:Tensor/2/float, inplace:bool/0/bool · returned Tensor · level recorded 2026-09-25
tinyimagenet_c_testloaders datasets/tiny_imagenet.py ran MIT (permissive) · 5f6874a404cdccdc · report fixture: signature name:str/0/str, serverity:str/0/str, batch_size:int/0/int, data_dir:str/0/str, num_workers:int/0/int, shuffle:bool/0/bool · returned DataLoader · level recorded 2026-09-25
tinyimagenet_dataloaders datasets/tiny_imagenet.py ran MIT (permissive) · 36e9a7bb8e92f93c · report fixture: signature batch_size:int/0/int, data_dir:str/0/str, num_workers:int/0/int, shuffle:bool/0/bool · returned tuple · level recorded 2026-09-25
cifar100_dataloaders in 8 papers datasets/cifar.py unverified MIT (permissive) · 2401e51688b0f4c4 · report
cifar10_dataloaders in 5 papers datasets/cifar.py unverified MIT (permissive) · 2d5dd53192074c7a · report
cifar10c_dataloaders datasets/cifar_c.py unverified MIT (permissive) · 6a51cdac8ae07f36 · report
linear_quant quantize_utils/func.py unverified MIT (permissive) · 9970e7f2fc273427 · report
resnet18 in 432 papers models/resnet.py unverified MIT (permissive) · 31501fc7e813d642 · report
tinyimagenet_c_trainloaders datasets/tiny_imagenet.py unverified MIT (permissive) · 9b92236376542dbd · report
train_epoch_adv trainer/engine_pgd.py unverified MIT (permissive) · 22d38cfde2b164df · report

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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