SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.01139 · 2026

-I ∈ R n×n : The Identity matrix; -A ∈ [0, 1] n×n : The Adjacency matrix of G; -D ∈ R n×n : The degree matrix, i.e., a diagonal matrix defined as D ii = ∑ n i=1 a ij ; -A ∈ R n×n : The normalized adjacency matrix of G, i.e., A = D -1/2 AD -1/2 ; -L ∈ R n×n : The Unnormalised Laplacian matrix of G; -Q ∈ R n×n : The Signless Laplacian matrix of G; -L rw ∈ R n×n : The Random-walk Normalised Laplacian of G; vii -L sym ∈ R n×n : The Symmetric Normalised Laplacian matrix of G; -Â ∈ R n×n : The Normalised Adjacency matrix of G; λ 1 (L rw ) : The Equidistribution radius of G; ϱ r (H) : The Normalized Spectral Gap of G; -X G ∈ R n×K , or simply X ∈ R n×K : The node feature matrix of the graph G, obtained by stacking up all x i vectors

L S T O F A C R O N Y M S A N D S Y M B O L S

arXiv · PDF

Code that ran

5 of 5 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 0 have no contract verdict; 4 failures of our instrument; 0 no recorded run), measured 2026-11-22; 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/2602.01139.json · badge: /api/ran/2602.01139.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
Sennadir/GCORN canonical (classifier label)
classifier basis: census
5 ran · 0 no recorded run
FunctionStatusWhere it lives
orthonormalize_weights Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Sennadir/GCORN/matrix_ortho.py
code served (permissive licence) · get_code("5c110b13ff423049")
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 (4, 3) 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

torch.manual_seed(0)

# Create a small weight matrix
# The function performs matrix operations: w.t().mm(w), so w should be a 2D tensor
# Using shape (4, 3) - a tall rectangular matrix suitable for orthonormalization
w = torch.randn(4, 3, dtype=torch.float32)

ARGS = {
    "w": w,
    "beta": 0.5,
    "iters": 20,
    "order": 3
}
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 (4, 3) — shape (4, 3) 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
parseval_weight_projections Ran
ran; our contract draft was wrong, not the code
Sennadir/GCORN/parseval_constraint.py
code served (permissive licence) · get_code("202ff7c0255b0b03")
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 SimpleModel. 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

torch.manual_seed(0)

class LinearModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.lin = nn.Linear(4, 3)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 4, kernel_size=3)
        self.conv2 = nn.Conv2d(4, 5, kernel_size=3)
        self.lin = LinearModule()
    
    def forward(self, x):
        return x

# Create model instance
model_temp = SimpleModel()

# Reshape weights to be 2D for matrix multiplication
# conv1.weight: (out_channels, in_channels, H, W) -> reshape to (out_channels, in_channels*H*W)
model_temp.conv1.weight.data = torch.randn(4, 27)
model_temp.conv2.weight.data = torch.randn(5, 36)
model_temp.lin.lin.weight.data = torch.randn(3, 4)

scale_param = 0.1
num_passes = 5

ARGS = {
    "model_temp": model_temp,
    "scale_param": scale_param,
    "num_passes": num_passes
}
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 SimpleModel — a SimpleModel output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModel 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
scale_values Ran
ran, honoured the contract we drafted
Sennadir/GCORN/matrix_ortho.py
code served (permissive licence) · get_code("bde2c248deee9c3f")
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 1, dtype family float, shape (1,) as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 numpy as np
import torch

torch.manual_seed(0)

# The function uses weight.shape[0] and weight.shape[1]
# So weight needs to be a tensor with at least 2 dimensions
weight = torch.randn(4, 3)

ARGS = {
    "weight": weight,
    "cuda": False,
}
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 1, dtype family float, shape (1,) — shape (1,) 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; our contract draft was wrong, not the code
Sennadir/GCORN/utils.py
code served (permissive licence) · get_code("fac5c5a475110d7a")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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

torch.manual_seed(0)

# Create a simple model that takes (x, adj) and outputs class logits
class SimpleGNNModel(nn.Module):
    def __init__(self, input_dim, hidden_dim, num_classes):
        super().__init__()
        self.linear1 = nn.Linear(input_dim, hidden_dim)
        self.linear2 = nn.Linear(hidden_dim, num_classes)
    
    def forward(self, x, adj):
        # x: (num_nodes, input_dim)
        # adj: (num_nodes, num_nodes)
        # Simple GNN: apply linear layers
        h = torch.relu(self.linear1(x))
        out = self.linear2(h)
        return out

# Create a simple data object with required attributes
class Data:
    def __init__(self):
        num_nodes = 6
        num_features = 4
        num_classes = 3
        
        # Node features
        self.x = torch.randn(num_nodes, num_features)
        
        # Node labels
        self.y = torch.tensor([0, 1, 2, 0, 1, 2])
        
        # Create masks for train/val/test
        self.train_mask = torch.tensor([True, True, False, False, False, False])
        self.val_mask = torch.tensor([False, False, True, True, False, False])
        self.test_mask = torch.tensor([False, False, False, False, True, True])

# Create model
num_nodes = 6
num_features = 4
num_classes = 3
model_local = SimpleGNNModel(num_features, 8, num_classes)
model_local.eval()

# Create data
data_local = Data()

# Create normalized adjacency matrix
adj_normalized = torch.eye(num_nodes) * 0.5 + torch.ones(num_nodes, num_nodes) * 0.1

ARGS = {
    "model_local": model_local,
    "data_local": data_local,
    "adj_normalized": adj_normalized,
}
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 list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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; our contract draft was wrong, not the code
Sennadir/GCORN/utils.py
code served (permissive licence) · get_code("2ee40e003ac849b9")
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 tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 takes x and adjacency as input
class SimpleGNN(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x, adj):
        # x: [num_nodes, input_dim]
        # adj: [num_nodes, num_nodes]
        # Simple forward: apply linear layer
        out = self.linear(x)
        return out

# Create model
model_local = SimpleGNN(input_dim=4, output_dim=3)

# Create optimizer
optimizer_local = optim.Adam(model_local.parameters(), lr=0.01)

# Create a minimal data object with required attributes
class Data:
    def __init__(self):
        # 2 nodes, 4 features
        self.x = torch.randn(2, 4)
        # 2 nodes, 3 classes
        self.y = torch.tensor([0, 1])
        # train_mask to select which nodes to train on
        self.train_mask = torch.tensor([True, False])

data_local = Data()

# Create normalized adjacency matrix [num_nodes, num_nodes]
adj_normalized = torch.eye(2)

ARGS = {
    "model_local": model_local,
    "optimizer_local": optimizer_local,
    "data_local": data_local,
    "adj_normalized": adj_normalized
}
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 tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
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

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 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 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.

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 Neural Networks (GNNs) have emerged as powerful tools for learning representations from structured data. Despite their growing popularity and success across various applications, GNNs encounter several challenges that limit their performance. in their generalization, robustness to adversarial perturbations, and the effectiveness of their representation learning capabilities. In this dissertation, I investigate these core aspects through three main contributions: (1) developing new representation learning techniques based on Graph Shift Operators (GSOs, aiming for enhanced performance across various contexts and applications, (2) introducing generalization-enhancing methods through graph data augmentation, and (3) developing more robust GNNs by leveraging orthonormalization techniques and noise-based defenses against adversarial attacks. By addressing these challenges, my work provides a more principled understanding of the limitations and potential of GNNs.

For agents

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

get_harvested_code_for_paper("2602.01139")
get_code_for_paper("2602.01139")
have("2602.01139")

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

curl https://syntology.ai/api/ran/2602.01139.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.01139.svg)](https://syntology.ai/paper/2602.01139)

Connect an agent — have() is free.