SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1905.11136 · 2019

Provably Powerful Graph Networks

arXiv · PDF · Open in the Atlas

Code that ran

4 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 0 failures of our instrument; 4 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/1905.11136.json · badge: /api/ran/1905.11136.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
hadarser/ProvablyPowerfulGraphNetworks_torch community (classifier label)
classifier basis not recorded on the link
4 ran · 3 no recorded run
hadarser/ProvablyPowerfulGraphNetworks reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
FunctionStatusWhere it lives
create_dirs Ran
ran, honoured the contract we drafted
hadarser/ProvablyPowerfulGraphNetworks_torch/utils/dirs.py
code served (permissive licence) · get_code("dbb32e482cdfa29f")
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 int, ndim 0, dtype family int, shape () 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 os
import tempfile
import shutil

# Create temporary directories for testing
temp_base = tempfile.mkdtemp()
test_dirs = [
    os.path.join(temp_base, 'dir1'),
    os.path.join(temp_base, 'dir2'),
    os.path.join(temp_base, 'nested', 'dir3')
]

ARGS = {
    'dirs': test_dirs
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, shape () — shape () 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
diag_offdiag_maxpool Ran
ran on a synthesized input
hadarser/ProvablyPowerfulGraphNetworks_torch/layers/layers.py
code served (permissive licence) · get_code("15bb4b29cca9b4a1")
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, 6) 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

torch.manual_seed(0)

# The function expects a tensor input with shape that allows:
# - input.shape[-1] to get N (last dimension size)
# - torch.diagonal(input, dim1=-2, dim2=-1) to work (needs at least 2D for the last two dims)
# - Indexing with dim=2 after diagonal, so needs at least 3 dimensions
# - Operations that preserve batch dimensions

# Looking at the function:
# - diagonal extracts diagonal across last 2 dims, result has shape [..., N]
# - max(..., dim=2)[0] suggests the result before max has at least 3 dims
# - The output concatenates on dim=1, suggesting (batch, channels, ...)
# - min_mat is reshaped to (1, 1, N, N), then subtracted from input
# - max operations on dim=3 and dim=2 suggest input is 4D: (batch, channels, N, N)

# So input should be: (batch_size, num_channels, N, N)
N = 4
batch_size = 2
num_channels = 3

input_tensor = torch.randn(batch_size, num_channels, N, N)

ARGS = {
    "input": input_tensor
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
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, 6) — shape (2, 6) 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
get_config_from_json Ran
ran on a synthesized input
hadarser/ProvablyPowerfulGraphNetworks_torch/utils/config.py
code served (permissive licence) · get_code("bbc60ae90dd59aa6")
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 EasyDict. 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 json
import tempfile
import os
from easydict import EasyDict

# Create a temporary JSON file with valid config data
temp_dir = tempfile.mkdtemp()
json_file_path = os.path.join(temp_dir, "config.json")

config_data = {
    "model_name": "test_model",
    "learning_rate": 0.001,
    "batch_size": 32,
    "epochs": 10,
    "optimizer": "adam",
    "layers": [64, 128, 256],
    "dropout": 0.5
}

with open(json_file_path, 'w') as f:
    json.dump(config_data, f)

ARGS = {
    "json_file": json_file_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
easydict — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EasyDict — a EasyDict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a EasyDict 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
process_config Ran
ran on a synthesized input
hadarser/ProvablyPowerfulGraphNetworks_torch/utils/config.py
code served (permissive licence) · get_code("a5ee96b6f2c405a2")
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 EasyDict. 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 json
import tempfile
import os
from easydict import EasyDict

# Create a temporary JSON config file
config_data = {
    "dataset_name": "MUTAG",
    "exp_name": "10fold_cross_validation",
    "target_param": False,
    "gpu": "0",
    "hyperparams": {
        "learning_rate": 0.001,
        "decay_rate": 0.5
    }
}

# Write config to a temporary file
temp_dir = tempfile.mkdtemp()
json_file = os.path.join(temp_dir, "config.json")
with open(json_file, 'w') as f:
    json.dump(config_data, f)

ARGS = {
    "json_file": json_file,
    "dataset_name": "MUTAG"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
easydict — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EasyDict — a EasyDict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a EasyDict 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
download_url Not yet run
unverified (harvested, no recorded run)
hadarser/ProvablyPowerfulGraphNetworks/utils/get_data.py
code served (permissive licence) · get_code("7838e53265948514")
How it ran

no invocation recorded.

load_dataset Not yet run
unverified (harvested, no recorded run)
hadarser/ProvablyPowerfulGraphNetworks_torch/data_loader/data_helper.py
code served (permissive licence) · get_code("9f1e79aa60bf79b8")
How it ran

no invocation recorded.

load_qm9 Not yet run
unverified (harvested, no recorded run)
hadarser/ProvablyPowerfulGraphNetworks_torch/data_loader/data_helper.py
code served (permissive licence) · get_code("5ffb58ac4245a0f8")
How it ran

no invocation recorded.

load_qm9_aux Not yet run
unverified (harvested, no recorded run)
hadarser/ProvablyPowerfulGraphNetworks_torch/data_loader/data_helper.py
code served (permissive licence) · get_code("6b4ad4ea61b1611e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 3 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 3 later) and 5 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.

FunctionAlso present inEarliest-datedOthers named
create_dirs 3 other papers
3 later
1907.10915 (2019-07)
Self-supervised Domain Adaptation for Computer Vision Tasks
2004.03990 (2020-04)
The general theory of permutation equivarant neural networks and higher order graph variational encoders
2410.11181 (2024-10)
DARNet: Dual Attention Refinement Network with Spatiotemporal Construction for Auditory Attention Detection
diag_offdiag_maxpool 2 other papers
2 later
2003.00982 (2020-03)
Benchmarking Graph Neural Networks
2106.06707 (2021-06)
Graph Neural Networks with Local Graph Parameters
get_config_from_json 1 other paper
1 later
2004.03990 (2020-04)
The general theory of permutation equivarant neural networks and higher order graph variational encoders

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

Recently, the Weisfeiler-Lehman (WL) graph isomorphism test was used to measure the expressive power of graph neural networks (GNN). It was shown that the popular message passing GNN cannot distinguish between graphs that are indistinguishable by the 1-WL test (Morris et al. 2018; Xu et al. 2019). Unfortunately, many simple instances of graphs are indistinguishable by the 1-WL test. In search for more expressive graph learning models we build upon the recent k-order invariant and equivariant graph neural networks (Maron et al. 2019a,b) and present two results: First, we show that such k-order networks can distinguish between non-isomorphic graphs as good as the k-WL tests, which are provably stronger than the 1-WL test for k>2. This makes these models strictly stronger than message passing models. Unfortunately, the higher expressiveness of these models comes with a computational cost of processing high order tensors. Second, setting our goal at building a provably stronger, simple and scalable model we show that a reduced 2-order network containing just scaled identity operator, augmented with a single quadratic operation (matrix multiplication) has a provable 3-WL expressive power. Differently put, we suggest a simple model that interleaves applications of standard Multilayer-Perceptron (MLP) applied to the feature dimension and matrix multiplication. We validate this model by presenting state of the art results on popular graph classification and regression tasks. To the best of our knowledge, this is the first practical invariant/equivariant model with guaranteed 3-WL expressiveness, strictly stronger than message passing models.

For agents

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

get_harvested_code_for_paper("1905.11136")
get_code_for_paper("1905.11136")
have("1905.11136")

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

curl https://syntology.ai/api/ran/1905.11136.json

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

[![Syntology run record](https://syntology.ai/api/ran/1905.11136.svg)](https://syntology.ai/paper/1905.11136)

Connect an agent — have() is free.