SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2210.07606 · NeurIPS · 2022

Revisiting Heterophily For Graph Neural Networks

Doina Precup, Sitao Luan, Mingde Zhao, Chenqing Hua, Xiao-Wen Chang, Shuyuan Zhang, Qincheng Lu, Jiaqi Zhu

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; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 2 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/2210.07606.json · badge: /api/ran/2210.07606.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
SitaoLuan/ACM-GNN canonical (classifier label)
classifier basis not recorded on the link
4 ran · 3 no recorded run
FunctionStatusWhere it lives
load_twitch_gamer Ran
ran; our contract draft was wrong, not the code
SitaoLuan/ACM-GNN/ACM-Geometric/load_data.py
pointer only (licence: MIT); the stored copy get_code serves, from jlinki/jdr, is pointer only (licence: NONE) · get_code("d4e6fd9f6cd8a3ad")
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 jlinki/jdr, 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_tensor Ran
ran on a synthesized input
SitaoLuan/ACM-GNN/ACM-Geometric/utils.py
code served (permissive licence) · get_code("3d6744ce51e1b5e3")
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 csr_matrix. 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 scipy.sparse as sp

# Create a small sparse matrix (4x3)
data = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
row = np.array([0, 1, 2, 2, 3])
col = np.array([0, 1, 0, 2, 1])
mx = sp.csr_matrix((data, (row, col)), shape=(4, 3))

ARGS = {
    "mx": mx,
    "eqvar": None
}
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 csr_matrix — a csr_matrix output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a csr_matrix 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
rand_train_test_idx Ran
ran on a synthesized input
SitaoLuan/ACM-GNN/ACM-Geometric/data_utils.py
code served (permissive licence) · get_code("e8000a86d21ce0e0")
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 torch
import numpy as np

torch.manual_seed(0)
np.random.seed(0)

# Create a label tensor with some labeled nodes (value != -1) and some unlabeled nodes (value == -1)
# Shape: (10,) with values 0-4 for labeled nodes and -1 for unlabeled nodes
label = torch.tensor([0, 1, 2, 3, 4, -1, -1, -1, -1, -1], dtype=torch.int64)

ARGS = {
    "label": label,
    "train_prop": 0.5,
    "valid_prop": 0.25,
    "ignore_negative": True,
}
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 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
sparse_mx_to_torch_sparse_tensor Ran
ran; our contract draft was wrong, not the code
SitaoLuan/ACM-GNN/ACM-Geometric/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from yimengmin/geometricscatteringmaximalclique, is pointer only (licence: NONE) · get_code("bdc1f1dad6d978e7")
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 yimengmin/geometricscatteringmaximalclique, 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.

even_quantile_labels Not yet run
unverified (harvested, no recorded run)
SitaoLuan/ACM-GNN/ACM-Geometric/data_utils.py
code served (permissive licence) · get_code("d305673ad3b29c2a")
How it ran

no invocation recorded.

load_fb100 Not yet run
unverified (harvested, no recorded run)
SitaoLuan/ACM-GNN/ACM-Geometric/load_data.py
code served (permissive licence) · get_code("a94acffdc2999ab9")
How it ran

no invocation recorded.

load_twitch Not yet run
unverified (harvested, no recorded run)
SitaoLuan/ACM-GNN/ACM-Geometric/load_data.py
code served (permissive licence) · get_code("166a96050e3dbbad")
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, 4 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 4 later) and 3 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
sparse_mx_to_torch_sparse_tensor 32 other papers
11 dated earlier, 21 later · 4 only as an assembly brick
1806.09055 (2018-06)
DARTS: Differentiable Architecture Search
1902.07153 (2019-02)
Simplifying Graph Convolutional Networks
2002.01633 (2020-02)
Structural Deep Clustering Network
and 29 more not named here
even_quantile_labels 8 other papers
8 later
2302.01524 (2023-02)
Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing
2306.08274 (2023-06)
A Simple and Scalable Graph Neural Network for Large Directed Graphs
2308.09517 (2023-08)
Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-based Similarity
and 5 more not named here
load_fb100 8 other papers
1 dated earlier, 7 later
2110.14446 (2021-10)
Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods
2302.01524 (2023-02)
Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing
2306.10759 (2023-06)
SGFormer: Simplifying and Empowering Transformers for Large-Graph Representations
and 5 more not named here
load_twitch_gamer 8 other papers
8 later
2302.01524 (2023-02)
Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing
2310.09202 (2023-10)
Graph Distillation with Eigenbasis Matching
2312.14438 (2023-12)
PC-Conv: Unifying Homophily and Heterophily with Two-fold Filtering
and 5 more not named here

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) extend basic Neural Networks (NNs) by using graph structures based on the relational inductive bias (homophily assumption). While GNNs have been commonly believed to outperform NNs in real-world tasks, recent work has identified a non-trivial set of datasets where their performance compared to NNs is not satisfactory. Heterophily has been considered the main cause of this empirical observation and numerous works have been put forward to address it. In this paper, we first revisit the widely used homophily metrics and point out that their consideration of only graph-label consistency is a shortcoming. Then, we study heterophily from the perspective of post-aggregation node similarity and define new homophily metrics, which are potentially advantageous compared to existing ones. Based on this investigation, we prove that some harmful cases of heterophily can be effectively addressed by local diversification operation. Then, we propose the Adaptive Channel Mixing (ACM), a framework to adaptively exploit aggregation, diversification and identity channels node-wisely to extract richer localized information for diverse node heterophily situations. ACM is more powerful than the commonly used uni-channel framework for node classification tasks on heterophilic graphs and is easy to be implemented in baseline GNN layers. When evaluated on 10 benchmark node classification tasks, ACM-augmented baselines consistently achieve significant performance gain, exceeding state-of-theart GNNs on most tasks without incurring significant computational burden.

For agents

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

get_harvested_code_for_paper("2210.07606")
get_code_for_paper("2210.07606")
have("2210.07606")

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

curl https://syntology.ai/api/ran/2210.07606.json

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

[![Syntology run record](https://syntology.ai/api/ran/2210.07606.svg)](https://syntology.ai/paper/2210.07606)

Connect an agent — have() is free.