SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.11198 · 2020

SIGN: Scalable Inception Graph Neural Networks

arXiv · PDF · Open in the Atlas

Code that ran

4 of 7 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; 2 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 3 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/2004.11198.json · badge: /api/ran/2004.11198.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
twitter-research/sign canonical (classifier label)
classifier basis: census
2 ran · 0 no recorded run
basiralab/falcon community (classifier label)
classifier basis not recorded on the link
2 ran · 3 no recorded run
FunctionStatusWhere it lives
contract_graph Ran
ran on a synthesized input
basiralab/falcon/contraction.py
code served (permissive licence) · get_code("6a91767a75ebfc9c")
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 Graph. 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 numpy as np
import networkx as nx
import random

# Create a small test graph with required node attributes
G = nx.Graph()

num_nodes = 10
num_features = 4
num_classes = 3

# Add nodes with required attributes
for i in range(num_nodes):
    x = np.random.randn(num_features)
    y = np.random.randint(0, num_classes, size=(num_classes,))
    train = i < 7  # Mark first 7 nodes as training nodes
    G.add_node(i, x=x, y=y, train=train)

# Add some edges to make it a connected graph
for i in range(num_nodes - 1):
    G.add_edge(i, i + 1)
    if i % 3 == 0:
        G.add_edge(i, (i + 2) % num_nodes)

ARGS = {
    'G': G,
    'num_features': num_features,
    'num_classes': num_classes,
    'centrality': 'EC',
    'constraint': 'FL',
    'N_cluster': 5,
    'gamma': 0.5,
    'steps': [5, 3],
    'multilabel': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
networkx, numpy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Graph — a Graph output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Graph output the recorded ndim 1 and dtype family int 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
precompute_SIGN_aggregation Ran
ran on a synthesized input
basiralab/falcon/models.py
code served (permissive licence) · get_code("51e2b6a3bef61bb0")
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, 16) 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)

# x: feature matrix, shape (num_nodes, num_features)
# adj: adjacency matrix, shape (num_nodes, num_nodes)
# max_hop: number of hops for aggregation

num_nodes = 2
num_features = 4

x = torch.randn(num_nodes, num_features)
adj = torch.randn(num_nodes, num_nodes)
max_hop = 3

ARGS = {
    "x": x,
    "adj": adj,
    "max_hop": max_hop
}
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, 16) — shape (2, 16) 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
test Ran
ran, honoured the contract we drafted
twitter-research/sign/sign_training.py
code served (permissive licence) · get_code("b40f6b945f2312d4")
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) and returned float, ndim 0, dtype family float, 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 torch
import torch.nn as nn
from tqdm import tqdm

torch.manual_seed(0)

# Define a minimal model that takes input and produces class logits
class SimpleClassifier(nn.Module):
    def __init__(self, input_dim=4, num_classes=3):
        super().__init__()
        self.linear = nn.Linear(input_dim, num_classes)
    
    def forward(self, x):
        return self.linear(x)

# Define a minimal evaluator with an eval method
class SimpleEvaluator:
    def eval(self, results):
        y_true = results['y_true']
        y_pred = results['y_pred']
        # Simple accuracy calculation
        acc = (y_true.squeeze() == y_pred.squeeze()).float().mean().item()
        return {'acc': acc}

# Create model
model = SimpleClassifier(input_dim=4, num_classes=3)

# Set device
device = torch.device('cpu')

# Create a simple data loader (list of (x, y) tuples)
# Batch size 2, input dim 4, 3 classes
loader = [
    (torch.randn(2, 4), torch.randint(0, 3, (2, 1))),
    (torch.randn(2, 4), torch.randint(0, 3, (2, 1))),
]

# Create evaluator
evaluator = SimpleEvaluator()

ARGS = {
    'model': model,
    'device': device,
    'loader': loader,
    'evaluator': evaluator,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, 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
train Ran
ran, honoured the contract we drafted
twitter-research/sign/sign_training.py
code served (permissive licence) · get_code("17c5046a153ae409")
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) and returned float, ndim 0, dtype family float, 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 torch
import torch.nn as nn
import torch.optim as optim

torch.manual_seed(0)

# Define a minimal model that satisfies the function's usage
class MinimalModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        # Returns log probabilities for nll_loss
        return F.log_softmax(self.linear(x), dim=1)

import torch.nn.functional as F

# Create model
model = MinimalModel()

# Set device
device = torch.device('cpu')

# Create a minimal train_loader as a list of (x, y) tuples
# x: batch of 2 samples, 4 features
# y: batch of 2 labels (class indices 0-2)
train_loader = [
    (torch.randn(2, 4), torch.randint(0, 3, (2, 1))),
    (torch.randn(2, 4), torch.randint(0, 3, (2, 1))),
]

# Add dataset attribute for len(train_loader.dataset)
class DataLoaderWrapper(list):
    def __init__(self, data):
        super().__init__(data)
        self.dataset = self
    
    def __len__(self):
        return sum(x.size(0) for x, y in self)

train_loader = DataLoaderWrapper(train_loader)

# Create optimizer
optimizer = optim.SGD(model.parameters(), lr=0.01)

ARGS = {
    'model': model,
    'device': device,
    'train_loader': train_loader,
    'optimizer': optimizer,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, 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
get_Organ Not yet run
unverified (harvested, no recorded run)
basiralab/falcon/datasets.py
code served (permissive licence) · get_code("cb5893426cdb115a")
How it ran

no invocation recorded.

get_PPI Not yet run
unverified (harvested, no recorded run)
basiralab/falcon/datasets.py
code served (permissive licence) · get_code("aca0d2c78e119514")
How it ran

no invocation recorded.

get_Planetoid Not yet run
unverified (harvested, no recorded run)
basiralab/falcon/datasets.py
code served (permissive licence) · get_code("313d854c25c7e5dd")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 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
test 1 other paper
1 later
2111.00064 (2021-11)
Node Feature Extraction by Self-Supervised Multi-scale Neighborhood Prediction
train 1 other paper
1 later
2111.00064 (2021-11)
Node Feature Extraction by Self-Supervised Multi-scale Neighborhood Prediction

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

Graph representation learning has recently been applied to a broad spectrum of problems ranging from computer graphics and chemistry to high energy physics and social media. The popularity of graph neural networks has sparked interest, both in academia and in industry, in developing methods that scale to very large graphs such as Facebook or Twitter social networks. In most of these approaches, the computational cost is alleviated by a sampling strategy retaining a subset of node neighbors or subgraphs at training time. In this paper we propose a new, efficient and scalable graph deep learning architecture which sidesteps the need for graph sampling by using graph convolutional filters of different size that are amenable to efficient precomputation, allowing extremely fast training and inference. Our architecture allows using different local graph operators (e.g. motif-induced adjacency matrices or Personalized Page Rank diffusion matrix) to best suit the task at hand. We conduct extensive experimental evaluation on various open benchmarks and show that our approach is competitive with other state-of-the-art architectures, while requiring a fraction of the training and inference time. Moreover, we obtain state-of-the-art results on ogbn-papers100M, the largest public graph dataset, with over 110 million nodes and 1.5 billion edges.

For agents

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

get_harvested_code_for_paper("2004.11198")
get_code_for_paper("2004.11198")
have("2004.11198")

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

curl https://syntology.ai/api/ran/2004.11198.json

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

[![Syntology run record](https://syntology.ai/api/ran/2004.11198.svg)](https://syntology.ai/paper/2004.11198)

Connect an agent — have() is free.