SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2306.05628 · 2023

Quantifying the Knowledge in GNNs for Reliable Distillation into MLPs

arXiv · PDF · Open in the Atlas

Code that ran

6 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; 0 honoured the contract we drafted, 2 violated it and 3 have no contract verdict; 1 failure of our instrument; 1 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/2306.05628.json · badge: /api/ran/2306.05628.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
lirongwu/rkd canonical (classifier label)
classifier basis: census
6 ran · 1 no recorded run
FunctionStatusWhere it lives
get_evaluator Ran
ran on a synthesized input
lirongwu/rkd/src/utils.py
code served (permissive licence) · get_code("ba8659b1d271a02a")
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 function. 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)

# The function get_evaluator takes a 'dataset' parameter and returns an evaluator function
# The evaluator function takes 'out' (model output) and 'labels' (ground truth labels)
# It computes accuracy by doing argmax(1) on out and comparing to labels
# So we need: dataset (can be anything, it's not used in the function body)
#            out: a tensor with shape (batch_size, num_classes)
#            labels: a tensor with shape (batch_size,) containing class indices

# Create a minimal dataset (not actually used by the function, but required as parameter)
dataset = None

ARGS = {
    "dataset": dataset
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
is_binary_bag_of_words Ran
ran, violated the contract we drafted
lirongwu/rkd/src/data_preprocess.py
pointer only (licence: MIT); the stored copy get_code serves, from YangLing0818/VQGraph, is pointer only (licence: NONE) · get_code("77c12b26c7311bbd")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from YangLing0818/VQGraph, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

normalize Ran
ran, violated the contract we drafted
lirongwu/rkd/src/data_preprocess.py
pointer only (licence: MIT); the stored copy get_code serves, from songw-sw/x-fnc, is pointer only (licence: NONE) · get_code("aa9c29936dee40a7")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from songw-sw/x-fnc, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

to_binary_bag_of_words Ran
ran; our contract draft was wrong, not the code
lirongwu/rkd/src/data_preprocess.py
pointer only (licence: MIT); the stored copy get_code serves, from xuChenSJTU/SAT-master-link-prediction, is pointer only (licence: NONE) · get_code("c1b14a7851940dca")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from xuChenSJTU/SAT-master-link-prediction, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

train Ran
ran on a synthesized input
lirongwu/rkd/src/train_and_eval.py
code served (permissive licence) · get_code("29faf1d5e00e0643")
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 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

torch.manual_seed(0)

class SimpleGNNModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, g, feats):
        return self.linear(feats)

model = SimpleGNNModel()
g = None

feats = torch.randn(2, 4)

labels = torch.tensor([0, 1])

criterion = nn.CrossEntropyLoss()

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

idx = torch.tensor([0, 1])

ARGS = {
    "model": model,
    "g": g,
    "feats": feats,
    "labels": labels,
    "criterion": criterion,
    "optimizer": optimizer,
    "idx": idx
}
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 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_mini_batch Ran
ran on a synthesized input
lirongwu/rkd/src/train_and_eval.py
code served (permissive licence) · get_code("b83839d3fe369101")
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

torch.manual_seed(0)

# Define a minimal model that accepts (None, feats) and outputs logits
class MinimalModel(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x, feats):
        # Function calls model(None, feats), so x is ignored
        return self.linear(feats)

# Create model
input_dim = 4
output_dim = 3
model = MinimalModel(input_dim, output_dim)

# Create small tensors
num_samples = 5
edge_idx = (torch.tensor([0, 1, 2]), torch.tensor([1, 2, 3]))
feats = torch.randn(num_samples, input_dim)
labels = torch.randint(0, output_dim, (num_samples,))
out_t_all = torch.randn(num_samples, output_dim)

# Index for training subset
idx = torch.tensor([0, 1, 2])

# Criterion functions
criterion_l = nn.NLLLoss()
criterion_t = nn.KLDivLoss(reduction='batchmean')

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

# Parameters dict
param = {
    'tau': 1.0,
    'lamb': 0.5
}

ARGS = {
    'model': model,
    'edge_idx': edge_idx,
    'feats': feats,
    'labels': labels,
    'out_t_all': out_t_all,
    'criterion_l': criterion_l,
    'criterion_t': criterion_t,
    'optimizer': optimizer,
    'idx': idx,
    'param': param
}
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
evaluate Not yet run
unverified (harvested, no recorded run)
lirongwu/rkd/src/train_and_eval.py
code served (permissive licence) · get_code("1901d498eb16daad")
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, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 3 later) and 4 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
normalize 24 other papers
18 dated earlier, 6 later
1609.02907 (2016-09)
Semi-Supervised Classification with Graph Convolutional Networks
1803.03735 (2018-03)
Attention-based Graph Neural Network for Semi-supervised Learning
1805.09112 (2018-05)
Hyperbolic Neural Networks
and 21 more not named here
to_binary_bag_of_words 4 other papers
3 dated earlier, 1 later
2006.16499 (2020-06)
SCE: Scalable Network Embedding from Sparsest Cut
2011.01623 (2020-11)
Learning on Attribute-Missing Graphs
2110.08727 (2021-10)
Graph-less Neural Networks: Teaching Old MLPs New Tricks via Distillation
and 1 more not named here
is_binary_bag_of_words 2 other papers
1 dated earlier, 1 later
2110.08727 (2021-10)
Graph-less Neural Networks: Teaching Old MLPs New Tricks via Distillation
2308.02117 (2023-08)
VQGraph: Rethinking Graph Representation Space for Bridging GNNs and MLPs

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

To bridge the gaps between topology-aware Graph Neural Networks (GNNs) and inference-efficient Multi-Layer Perceptron (MLPs), GLNN proposes to distill knowledge from a well-trained teacher GNN into a student MLP. Despite their great progress, comparatively little work has been done to explore the reliability of different knowledge points (nodes) in GNNs, especially their roles played during distillation. In this paper, we first quantify the knowledge reliability in GNN by measuring the invariance of their information entropy to noise perturbations, from which we observe that different knowledge points (1) show different distillation speeds (temporally); (2) are differentially distributed in the graph (spatially). To achieve reliable distillation, we propose an effective approach, namely Knowledge-inspired Reliable Distillation (KRD), that models the probability of each node being an informative and reliable knowledge point, based on which we sample a set of additional reliable knowledge points as supervision for training student MLPs. Extensive experiments show that KRD improves over the vanilla MLPs by 12.62% and outperforms its corresponding teacher GNNs by 2.16% averaged over 7 datasets and 3 GNN architectures.

For agents

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

get_harvested_code_for_paper("2306.05628")
get_code_for_paper("2306.05628")
have("2306.05628")

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

curl https://syntology.ai/api/ran/2306.05628.json

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

[![Syntology run record](https://syntology.ai/api/ran/2306.05628.svg)](https://syntology.ai/paper/2306.05628)

Connect an agent — have() is free.