Papers › Preventing Gradient Attenuation in Lipschitz Constrained Convolutional Networks

Preventing Gradient Attenuation in Lipschitz Constrained Convolutional Networks

3 Nov 2019NeurIPS 2019 12arXiv:1911.00937archive 2025-07-28

Qiyang Li, Saminul Haque, Cem Anil, James Lucas, Roger Grosse, Jörn-Henrik Jacobsen

Lipschitz constraints under L2 norm on deep neural networks are useful for provable adversarial robustness bounds, stable training, and Wasserstein distance estimation. While heuristic approaches such as the gradient penalty have seen much practical success, it is challenging to achieve similar practical performance while provably enforcing a Lipschitz constraint. In principle, one can design Lipschitz constrained architectures using the composition property of Lipschitz functions, but Anil et al. recently identified a key obstacle to this approach: gradient norm attenuation. They showed how to circumvent this problem in the case of fully connected networks by designing each layer to be gradient norm preserving. We extend their approach to train scalable, expressive, provably Lipschitz convolutional networks. In particular, we present the Block Convolution Orthogonal Parameterization (BCOP), an expressive parameterization of orthogonal convolution operations. We show that even though the space of orthogonal convolutions is disconnected, the largest connected component of BCOP with 2n channels can represent arbitrary BCOP convolutions over n channels. Our BCOP parameterization allows us to train large convolutional networks with provable Lipschitz bounds. Empirically, we find that it is competitive with existing approaches to provable adversarial robustness and Wasserstein distance estimation.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

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 14 code samples harvested from 1 repository linked to this paper: 10 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 2 honoured, 0 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 4 unverified. Where Syntology's instrument failed: 4 where our contract draft was wrong; 4 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 14 samples from 1 repository, 10 ran (2 with no instrument failure; 8 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 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

ColinQiyangLi/LConvNet officialmentioned in papermentioned on GitHubpytorchSyntology: 14 samples, 10 ran (8 instrument-failed) · 4 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

14 samples harvested: 10 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 2 honoured, 0 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 4 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.

2ran · honoured contract
4ran · our draft was wrong
4ran; the contract check could not run
4unverified

Licence: 0 of the 14 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 ColinQiyangLi/LConvNet. “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.

ColinQiyangLi/LConvNet official repository 14 samples, 10 ran (8 instrument-failed) · 4 unverified
bjorck_orthonormalize lconvnet/layers/bcop.py ran; the contract check could not run (RAISES) fingerprinted MIT (permissive) · c6d2da76d5fbb78a · report fixture: signature w:Tensor/3/float, beta:float/0/float, iters:int/0/int, order:int/0/int, power_iteration_scaling:bool/0/bool, default_scaling:bool/0/bool · returned Tensor · level recorded 2026-09-22
block_orth lconvnet/layers/bcop.py ran · our draft was wrong MIT (permissive) · 67faa492a8df483a · report fixture: signature p1:Tensor/2/float, p2:Tensor/2/float · returned dict · level recorded 2026-09-22
conv2d_cyclic_pad lconvnet/layers/bcop.py ran · our draft was wrong fingerprinted MIT (permissive) · 0e56ee61b3f7a88d · report fixture: signature x:Tensor/4/float, weight:Tensor/4/float, bias:Tensor/1/float · returned Tensor · level recorded 2026-09-22
conv_singular_values_numpy lconvnet/layers/bcop.py ran; the contract check could not run (RAISES) MIT (permissive) · 8c6d22a2cf2995bc · report fixture: signature kernel:ndarray/4/float, input_shape:tuple/1/int · returned ndarray · level recorded 2026-09-22
cyclic_pad_2d lconvnet/layers/bcop.py ran; the contract check could not run (RAISES) fingerprinted MIT (permissive) · a03b76e7e8b7145d · report fixture: signature x:Tensor/4/float, pads:tuple/1/int, even_h:bool/0/bool, even_w:bool/0/bool · returned Tensor · level recorded 2026-09-22
dict_to_tensor lconvnet/layers/bcop.py ran · honoured contract MIT (permissive) · 2bdb69b2570e1bd8 · report fixture: signature x:dict/0/?, k1:int/0/int, k2:int/0/int · returned Tensor · level recorded 2026-09-24
matrix_conv lconvnet/layers/bcop.py ran · our draft was wrong MIT (permissive) · 20d0aab322225a36 · report fixture: signature m1:dict/0/?, m2:dict/0/? · returned dict · level recorded 2026-09-22
orthogonal_matrix lconvnet/layers/bcop.py ran · honoured contract fingerprinted MIT (permissive) · 6e1165eb18d80570 · report fixture: signature n:int/0/int · returned Tensor · level recorded 2026-09-22
power_iteration in 30 papers lconvnet/layers/bcop.py ran; the contract check could not run (RAISES) fingerprinted MIT (permissive) · 264dacf0aae30538 · report fixture: signature A:Tensor/2/float, init_u:Tensor/2/float, n_iters:int/0/int, return_uv:bool/0/bool · returned tuple · level recorded 2026-09-22
symmetric_projection lconvnet/layers/bcop.py ran · our draft was wrong fingerprinted MIT (permissive) · 246c26b7c160c814 · report fixture: signature n:int/0/int, ortho_matrix:Tensor/2/float, mask:Tensor/1/float · returned Tensor · level recorded 2026-09-22
BCOP lconvnet/layers/bcop.py unverified MIT (permissive) · 568d86e4f1f7b544 · report
LipschitzModuleL2 lconvnet/layers/bcop.py unverified MIT (permissive) · b3df67d1d29dafde · report
StreamlinedModule lconvnet/layers/bcop.py unverified MIT (permissive) · 33d734483d94ad69 · report
convolution_orthogonal_generator_projs lconvnet/layers/bcop.py unverified MIT (permissive) · 5a14b8c3f447240f · 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 BCOP (graph method: BCOP). 12 bricks; individually on a fixture: 8 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 7 where Syntology's instrument failed) · 4 unverified. The assembly has not been run end-to-end.

  1. BCOP entry · level 0 · unverified · 568d86e4f1f7b544
  2. LipschitzModuleL2 helper · level 0 · unverified · b3df67d1d29dafde
  3. StreamlinedModule helper · level 0 · unverified · 33d734483d94ad69
  4. bjorck_orthonormalize helper · level 1 · ran; the contract check could not run (RAISES) · c6d2da76d5fbb78a
  5. conv2d_cyclic_pad helper · level 1 · ran · our draft was wrong · 0e56ee61b3f7a88d
  6. conv_singular_values_numpy helper · level 1 · ran; the contract check could not run (RAISES) · 8c6d22a2cf2995bc
  7. convolution_orthogonal_generator_projs helper · level 0 · unverified · 5a14b8c3f447240f
  8. power_iteration helper · level 1 · ran; the contract check could not run (RAISES) · 264dacf0aae30538
  9. cyclic_pad_2d helper · level 1 · ran; the contract check could not run (RAISES) · a03b76e7e8b7145d
  10. block_orth helper · level 1 · ran · our draft was wrong · 67faa492a8df483a
  11. dict_to_tensor helper · level 1 · ran · honoured contract · 2bdb69b2570e1bd8
  12. matrix_conv helper · level 1 · ran · our draft was wrong · 20d0aab322225a36

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="1911.00937") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Adversarial Robustness

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Convolution

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