SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.03169 · ICLR · 2023

A Unified Algebraic Perspective on Lipschitz Neural Networks

Bin Hu, Alexandre Allauzen, Alexandre Araujo, Aaron Havens, Blaise Delattre

arXiv · PDF · Open in the Atlas

Code that ran

2 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 1 has no contract verdict; 0 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/2303.03169.json · badge: /api/ran/2303.03169.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
araujoalexandre/lipschitz-sll-networks canonical (classifier label)
classifier basis: classifier_2025
2 ran · 2 no recorded run
FunctionStatusWhere it lives
SDPBasedLipschitzConvLayer Ran
ran on a synthesized input · constructed an object; no result was computed
araujoalexandre/lipschitz-sll-networks/core/models/layers.py
code served (permissive licence) · get_code("a53f2e98004de3e5")
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 an instance of its own class, SDPBasedLipschitzConvLayer: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type 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
import torch.nn as nn
import numpy as np

torch.manual_seed(0)

# Input tensor for forward pass: batch_size=2, channels=4, height=8, width=8
x = torch.randn(2, 4, 8, 8)

# Arguments for SDPBasedLipschitzConvLayer
ARGS = {
    'cin': 4,           # input channels
    'inner_dim': 5,     # inner dimension (distinct from cin)
    'kernel_size': 3,   # kernel size
    'stride': 1         # stride
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SDPBasedLipschitzConvLayer — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a SDPBasedLipschitzConvLayer output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
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
safe_inv Ran
ran, violated the contract we drafted
araujoalexandre/lipschitz-sll-networks/core/models/layers.py
code served (permissive licence) · get_code("78b329790b439bbe")
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 2, dtype family float, shape (2, 4) 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
import numpy as np

torch.manual_seed(0)

# safe_inv takes a tensor x and performs element-wise inverse with masking
# It checks for zero values, computes x^(-1), and sets inverse of zeros to 0

x = torch.tensor([[1.0, 2.0, 0.0, 4.0],
                   [5.0, 0.0, 3.0, 2.0]], dtype=torch.float32)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
eval_certified Not yet run
unverified (harvested, no recorded run)
araujoalexandre/lipschitz-sll-networks/reproduce_tables.py
code served (permissive licence) · get_code("176682a411532153")
How it ran

no invocation recorded.

safe_inv Not yet run
unverified (harvested, no recorded run)
araujoalexandre/lipschitz-sll-networks/core/models/layers.py
code served (permissive licence) · get_code("5da48aeb96343d0a")
How it ran

no invocation recorded. this sample carries JSON test cases instead of a setup block; get_reference_implementation serves them with the level they back.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 3 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.

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

Important research efforts have focused on the design and training of neural networks with a controlled Lipschitz constant. The goal is to increase and sometimes guarantee the robustness against adversarial attacks. Recent promising techniques draw inspirations from different backgrounds to design 1-Lipschitz neural networks, just to name a few: convex potential layers derive from the discretization of continuous dynamical systems, Almost-Orthogonal-Layer proposes a tailored method for matrix rescaling. However, it is today important to consider the recent and promising contributions in the field under a common theoretical lens to better design new and improved layers. This paper introduces a novel algebraic perspective unifying various types of 1-Lipschitz neural networks, including the ones previously mentioned, along with methods based on orthogonality and spectral methods. Interestingly, we show that many existing techniques can be derived and generalized via finding analytical solutions of a common semidefinite programming (SDP) condition. We also prove that AOL biases the scaled weight to the ones which are close to the set of orthogonal matrices in a certain mathematical manner. Moreover, our algebraic condition, combined with the Gershgorin circle theorem, readily leads to new and diverse parameterizations for 1-Lipschitz network layers. Our approach, called SDP-based Lipschitz Layers (SLL), allows us to design non-trivial yet efficient generalization of convex potential layers. Finally, the comprehensive set of experiments on image classification shows that SLLs outperform previous approaches on certified robust accuracy. Code is available at github.com/araujoalexandre/Lipschitz-SLL-Networks. (10/26/2023): Erratum is added in Appendix D. This is an updated version that fixes an implementation issue in the previous version. Due to that implementation issue, the original numerical results in our original ICLR paper are not accurate. We elaborate on the issue and provides some fix in Appendix D.

For agents

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

get_harvested_code_for_paper("2303.03169")
get_code_for_paper("2303.03169")
have("2303.03169")

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

curl https://syntology.ai/api/ran/2303.03169.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.03169.svg)](https://syntology.ai/paper/2303.03169)

Connect an agent — have() is free.