SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2001.06216 · 2020

GraphLIME: Local Interpretable Model Explanations for Graph Neural Networks

arXiv · PDF · Open in the Atlas

Code that ran

9 of 13 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; 0 honoured the contract we drafted, 0 violated it and 9 have no contract verdict; 0 failures of our instrument; 4 no recorded run), measured 2026-11-21; 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/2001.06216.json · badge: /api/ran/2001.06216.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
riken-aip/pyHSICLasso community (classifier label)
classifier basis not recorded on the link
7 ran · 3 no recorded run
WilliamCCHuang/GraphLIME community (classifier label)
classifier basis not recorded on the link
2 ran · 1 no recorded run
FunctionStatusWhere it lives
add_noise_features Ran
ran on a synthesized input
WilliamCCHuang/GraphLIME/exp/noise_features/utils.py
code served (permissive licence) · get_code("3b0362788eafb47b")
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 MockData. 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

torch.manual_seed(0)

class MockData:
    def __init__(self):
        self.x = torch.randn((2, 4))

data = MockData()
num_noise = 3

ARGS = {
    "data": data,
    "num_noise": num_noise,
}
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 MockData — a MockData output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MockData 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
input_csv_file Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/input_data.py
code served (permissive licence) · get_code("76b2ea7cba9b77a9")
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. 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 pandas as pd
import tempfile
import os

# Create a temporary CSV file for testing
temp_dir = tempfile.mkdtemp()
csv_file = os.path.join(temp_dir, "test_data.csv")

# Create sample data with multiple columns
data = {
    'feature1': [1.0, 2.0, 3.0],
    'feature2': [4.0, 5.0, 6.0],
    'feature3': [7.0, 8.0, 9.0],
    'class': [0, 1, 0]
}
df = pd.DataFrame(data)
df.to_csv(csv_file, index=False, sep=',')

ARGS = {
    'file_name': csv_file,
    'output_list': ['class']
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is 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
input_tsv_file Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/input_data.py
code served (permissive licence) · get_code("06b2bab851a15ffa")
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. 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 pandas as pd
import tempfile
import os

# Create a temporary CSV file with test data
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test_data.tsv")

# Create a small TSV file with sample data
data = {
    'feature1': [1.0, 2.0],
    'feature2': [3.0, 4.0],
    'feature3': [5.0, 6.0],
    'class': [0, 1]
}
df = pd.DataFrame(data)
df.to_csv(temp_file, sep='\t', index=False)

ARGS = {
    'file_name': temp_file,
    'output_list': ['class']
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is 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
input_txt_file Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/input_data.py
code served (permissive licence) · get_code("10d789c9198d106b")
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. 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 pandas as pd
import tempfile
import os

# Create a temporary CSV file with test data
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test_data.csv")

# Create a simple CSV file with some features and outputs
data = {
    'feature1': [1.0, 2.0, 3.0, 4.0],
    'feature2': [5.0, 6.0, 7.0, 8.0],
    'feature3': [9.0, 10.0, 11.0, 12.0],
    'target1': [0.5, 1.5, 2.5, 3.5],
    'target2': [10.0, 20.0, 30.0, 40.0]
}

df = pd.DataFrame(data)
df.to_csv(temp_file, sep=',', index=False)

ARGS = {
    'file_name': temp_file,
    'output_list': ['target1', 'target2'],
    'sep': ','
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is 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
kernel_delta Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/kernel_tools.py
code served (permissive licence) · get_code("ce1a4b24283c1723")
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 ndarray, ndim 2, dtype family float, shape (6, 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 numpy as np

# Create two simple test arrays with shape (1, n) where n is the number of columns
# The function checks X_in_1.shape[1] and X_in_2.shape[1]
# and uses np.where to find indices matching unique values

X_in_1 = np.array([[1, 2, 1, 3, 2, 1]])  # shape (1, 6)
X_in_2 = np.array([[1, 3, 2, 1, 2, 3]])  # shape (1, 6)

ARGS = {
    "X_in_1": X_in_1,
    "X_in_2": X_in_2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (6, 6) — shape (6, 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
kernel_delta_norm Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/kernel_tools.py
code served (permissive licence) · get_code("9c39eb068955fe68")
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 ndarray, 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 numpy as np

# Construct valid arguments for kernel_delta_norm
# The function expects two numpy arrays where:
# - shape[1] gives the number of columns
# - The arrays contain discrete values (since it uses == comparisons)
# - It counts occurrences of each unique value

# Create small test arrays with repeated values
X_in_1 = np.array([
    [1, 2, 1, 2],
    [1, 1, 2, 2],
    [1, 2, 1, 1]
])  # shape: (3, 4) - 3 rows, 4 columns

X_in_2 = np.array([
    [1, 2, 1],
    [2, 1, 2],
    [1, 1, 2]
])  # shape: (3, 3) - 3 rows, 3 columns

ARGS = {
    "X_in_1": X_in_1,
    "X_in_2": X_in_2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, 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
kernel_gaussian Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/kernel_tools.py
code served (permissive licence) · get_code("1c45fa0ff4d5fc3f")
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 ndarray, ndim 2, dtype family float, shape (3, 2) 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 numpy as np

# Create small numpy arrays for X_in_1 and X_in_2
# X_in_1 should have shape (d, n_1) and X_in_2 should have shape (d, n_2)
# Using distinct sizes: 4 features, 3 samples for X_in_1, 2 samples for X_in_2
X_in_1 = np.random.randn(4, 3)
X_in_2 = np.random.randn(4, 2)

# Sigma parameter for the Gaussian kernel
sigma = 1.0

ARGS = {
    "X_in_1": X_in_1,
    "X_in_2": X_in_2,
    "sigma": sigma
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (3, 2) — shape (3, 2) 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
modify_train_mask Ran
ran on a synthesized input
WilliamCCHuang/GraphLIME/exp/noise_features/utils.py
code served (permissive licence) · get_code("1bf9d9934a3aafd0")
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 Data. 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 numpy as np

torch.manual_seed(0)

class Data:
    def __init__(self):
        self.x = torch.randn(10, 4)
        self.train_mask = torch.zeros(10, dtype=torch.bool)
        self.val_mask = torch.zeros(10, dtype=torch.bool)
        self.test_mask = torch.zeros(10, dtype=torch.bool)

data = Data()

ARGS = {
    "data": data
}
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 Data — a Data output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Data 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
nlars Ran
ran on a synthesized input
riken-aip/pyHSICLasso/pyHSICLasso/nlars.py
code served (permissive licence) · get_code("beb23901f71ce336")
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. 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
from scipy.sparse import lil_matrix

# Set random seed for reproducibility
np.random.seed(0)

# Create small test data
# X should be D x D matrix
D = 5
X = np.random.randn(D, D).astype(np.float32)

# X_ty should be D x 1 vector
X_ty = np.random.randn(D, 1).astype(np.float32)

# num_feat: number of features to extract
num_feat = 3

# max_neighbors: maximum number of neighbors to track
max_neighbors = 2

ARGS = {
    'X': X,
    'X_ty': X_ty,
    'num_feat': num_feat,
    'max_neighbors': max_neighbors
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is 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
compute_kernel Not yet run
unverified (harvested, no recorded run)
riken-aip/pyHSICLasso/pyHSICLasso/hsic_lasso.py
code served (permissive licence) · get_code("5690b8a912e41025")
How it ran

no invocation recorded.

hsic_lasso Not yet run
unverified (harvested, no recorded run)
riken-aip/pyHSICLasso/pyHSICLasso/hsic_lasso.py
code served (permissive licence) · get_code("aaa3f8bb4257d737")
How it ran

no invocation recorded.

parallel_compute_kernel Not yet run
unverified (harvested, no recorded run)
riken-aip/pyHSICLasso/pyHSICLasso/hsic_lasso.py
code served (permissive licence) · get_code("d06b223d2d62ef20")
How it ran

no invocation recorded.

prepare_data Not yet run
unverified (harvested, no recorded run)
WilliamCCHuang/GraphLIME/exp/noise_features/utils.py
code served (permissive licence) · get_code("b6860453fe46f024")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 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 13 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 2026-11-21.

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 structured data has wide applicability in various domains such as physics, chemistry, biology, computer vision, and social networks, to name a few. Recently, graph neural networks (GNN) were shown to be successful in effectively representing graph structured data because of their good performance and generalization ability. GNN is a deep learning based method that learns a node representation by combining specific nodes and the structural/topological information of a graph. However, like other deep models, explaining the effectiveness of GNN models is a challenging task because of the complex nonlinear transformations made over the iterations. In this paper, we propose GraphLIME, a local interpretable model explanation for graphs using the Hilbert-Schmidt Independence Criterion (HSIC) Lasso, which is a nonlinear feature selection method. GraphLIME is a generic GNN-model explanation framework that learns a nonlinear interpretable model locally in the subgraph of the node being explained. More specifically, to explain a node, we generate a nonlinear interpretable model from its $N$-hop neighborhood and then compute the K most representative features as the explanations of its prediction using HSIC Lasso. Through experiments on two real-world datasets, the explanations of GraphLIME are found to be of extraordinary degree and more descriptive in comparison to the existing explanation methods.

For agents

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

get_harvested_code_for_paper("2001.06216")
get_code_for_paper("2001.06216")
have("2001.06216")

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

curl https://syntology.ai/api/ran/2001.06216.json

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

[![Syntology run record](https://syntology.ai/api/ran/2001.06216.svg)](https://syntology.ai/paper/2001.06216)

Connect an agent — have() is free.