Papers › Robust CLIP: Unsupervised Adversarial Fine-Tuning of Vision Embeddings for Robust...
Robust CLIP: Unsupervised Adversarial Fine-Tuning of Vision Embeddings for Robust Large Vision-Language Models
Christian Schlarmann, Naman Deep Singh, Francesco Croce, Matthias Hein
Multi-modal foundation models like OpenFlamingo, LLaVA, and GPT-4 are increasingly used for various real-world tasks. Prior work has shown that these models are highly vulnerable to adversarial attacks on the vision modality. These attacks can be leveraged to spread fake information or defraud users, and thus pose a significant risk, which makes the robustness of large multi-modal foundation models a pressing problem. The CLIP model, or one of its variants, is used as a frozen vision encoder in many large vision-language models (LVLMs), e.g. LLaVA and OpenFlamingo. We propose an unsupervised adversarial fine-tuning scheme to obtain a robust CLIP vision encoder, which yields robustness on all vision down-stream tasks (LVLMs, zero-shot classification) that rely on CLIP. In particular, we show that stealth-attacks on users of LVLMs by a malicious third party providing manipulated images are no longer possible once one replaces the original CLIP model with our robust one. No retraining or fine-tuning of the down-stream LVLMs is required. The code and robust models are available at https://github.com/chs20/RobustVLM
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):
get_harvested_code_for_paper(arxiv_id="2402.12336")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2402.12336", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2402.12336")
Code
Syntology 9 code samples harvested from 1 repository linked to this paper: 7 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 6 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; 3 where the contract check could not run; failures of Syntology's instrument, not of the code.
By repository: official repository: 9 samples from 1 repository, 7 ran (1 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 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 4 of the 7 samples that ran have no recorded date.
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
9 samples harvested: 7 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 6 where Syntology's instrument failed) · 2 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.
Licence: 0 of the 9 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.
Harvested from chs20/robustvlm. “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.
5bfccc77a86a50fb · report
fixture: signature embedding:Tensor/2/float, targets:Tensor/1/int · returned Tensor · level recorded 2026-09-22 census: 3 implementations from 5 papers, 3 distinct outputs across 3 buckets, one shared input per bucket train/adversarial_training_clip.py ran · our draft was wrong MIT (permissive) ·
1f2be2216dc3f7cd · report
fixture: signature logits:Tensor/2/float, targets:Tensor/1/int · returned float · level recorded, date not recorded census: 11 implementations from 11 papers, 9 distinct outputs across 7 buckets, one shared input per bucket this sample: bucket 5 of 7, arg 1: rank 2, kind float, dtype float32 · arg 2: rank 1, kind int, dtype int64 · arg 3: rank 2, kind float, dtype float32 · arg 4: rank 0, kind float, dtype float32 train/adversarial_training_clip.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant MIT (permissive) ·
75f68885e2ff2684 · report
fixture: signature loss_str:str/0/str, embedding:Tensor/2/float, targets:Tensor/1/int, embedding_orig:Tensor/2/float, logit_scale:Tensor/0/float, embedding_text_labels_norm:NoneType/0/?, reduction:str/0/str · returned Tensor · level recorded, date not recorded census: 11 implementations from 11 papers, 9 distinct outputs across 7 buckets, one shared input per bucket this sample: bucket 1 of 7, arg 1: rank 2, kind float, dtype float32 · arg 2: rank 1, kind int, dtype int64 · arg 3: rank 2, kind float, dtype float32 · arg 4: rank 0, kind float, dtype float32 · arg 5: rank 2, kind float, dtype float32 train/adversarial_training_clip.py ran; the contract check could not run (RAISES) MIT (permissive) ·
b03f0274dabbb8c0 · report
fixture: signature loss_str:str/0/str, embedding:Tensor/2/float, targets:Tensor/1/int, embedding_orig:Tensor/2/float, logit_scale:Tensor/0/float, embedding_text_labels_norm:Tensor/2/float, reduction:str/0/str · returned Tensor · level recorded 2026-09-22 68d0f41190a26308 · report
fixture: signature train_dataset:SimpleDataset/0/?, query_set_size:int/0/int, seed:int/0/int · returned list · level recorded, date not recorded 63ec736a72a6cb74 · report
fixture: signature num_samples:int/0/int, query_set_size:int/0/int, full_dataset:list/1/int, seed:int/0/int · returned ndarray · level recorded, date not recorded 7062c33b2cc96026 · report
fixture: signature forward:callable/0/?, loss_fn:callable/0/?, data_clean:Tensor/2/float, targets:Tensor/1/int, norm:str/0/str, eps:float/0/float, iterations:int/0/int, stepsize:float/0/float, output_normalize:bool/0/bool, perturbation:NoneType/0/?, mode:str/0/str, momentum:float/0/float, verbose:bool/0/bool · returned Tensor · level recorded 2026-09-22 1707d13b9dc6cb32 · report
04cf221e803c96d7 · report
Method as code Syntology
What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.
1 assembly on this paper: 1 harvested, 0 generated.
harvested Assembly for Robust CLIP (graph method: Robust CLIP). 15 bricks; individually on a fixture: 13 ran (of which 1 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 10 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.
- train_one_epoch entry · level 0 · unverified ·
1707d13b9dc6cb32 - AverageMeter helper · level 1 · ran ·
52b386c85c15f1f1 - ComputeLossWrapper loss · level 1 · ran ·
5bfccc77a86a50fb - compute_acc metric · level 1 · ran · our draft was wrong ·
1f2be2216dc3f7cd - compute_loss loss · level 1 · ran; the contract check could not run (RAISES) ·
b03f0274dabbb8c0 - pgd helper · level 1 · ran · our draft was wrong ·
7062c33b2cc96026 - unwrap_model helper · level 0 · unverified ·
04cf221e803c96d7 - ce loss · level 1 · ran; the contract check could not run ·
7ebb8ba5f73a84b6 - l2 loss · level 1 · ran · our draft was wrong ·
df3d5eb6fd2523a1 - normalize_grad helper · level 1 · ran; the contract check could not run ·
6fffb629c7f9b304 - project_perturbation helper · level 1 · ran; the contract check could not run ·
74d23c944bf73c1e - L0_norm helper · level 1 · ran · honoured contract ·
a1b32de808e46be1 - L1_projection helper · level 1 · ran · our draft was wrong ·
f0f026ce0c0e2f39 - L2_norm helper · level 1 · ran; the contract check could not run ·
fa19820305aabe1d - check_oscillation helper · level 1 · ran · our draft was wrong ·
1b13d7d6e5549507
Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2402.12336") returns these assemblies with each brick's role and wiring (how to connect).
Tasks
1 archive task tag without a task page not shown.
Results from the paper archive 2025-07-28
No leaderboard rows for this paper in the archive.
Methods
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