Papers › Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs

Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs

20 Jun 2020NeurIPS 2020 12arXiv:2006.11468archive 2025-07-28

Jiong Zhu, Yujun Yan, Lingxiao Zhao, Mark Heimann, Leman Akoglu, Danai Koutra

We investigate the representation power of graph neural networks in the semi-supervised node classification task under heterophily or low homophily, i.e., in networks where connected nodes may have different class labels and dissimilar features. Many popular GNNs fail to generalize to this setting, and are even outperformed by models that ignore the graph structure (e.g., multilayer perceptrons). Motivated by this limitation, we identify a set of key designs -- ego- and neighbor-embedding separation, higher-order neighborhoods, and combination of intermediate representations -- that boost learning from the graph structure under heterophily. We combine them into a graph neural network, H2GCN, which we use as the base method to empirically evaluate the effectiveness of the identified designs. Going beyond the traditional benchmarks with strong homophily, our empirical analysis shows that the identified designs increase the accuracy of GNNs by up to 40% and 27% over models without them on synthetic and real networks with heterophily, respectively, and yield competitive performance under homophily.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Syntology holds a reference implementation of H2GCN, generated from the paper text (not the authors' code), at level 3 on Syntology's ladder: V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)”; V3 “paper-derived property tests pass on both of those same two generations.” (quoted from Syntology's server card). No rung compares this code against the authors' implementation. For agents: get_reference_implementation(name='H2GCN') (how to connect); the free check have('2006.11468') needs no token.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 21 code samples harvested from 5 repositories linked to this paper: 16 ran (of which 2 constructed an object rather than computing a result; 13 with no instrument failure: 4 honoured, 2 violated, 7 with no contract checked; 3 where Syntology's instrument failed) · 5 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 4 samples from 1 repository, 3 ran (2 with no instrument failure; 1 where Syntology's instrument failed); community (archive-listed): 1 sample from 1 repository, 1 ran (1 with no instrument failure; 0 where Syntology's instrument failed); found in paper text by Syntology: 16 samples from 3 repositories, 12 ran (10 with no instrument failure; 2 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 16 samples that ran has no recorded date.

GemsLab/H2GCN officialmentioned in papermentioned on GitHubtfSyntology: 4 samples, 3 ran (1 instrument-failed) · 1 unverified report
GitEventhandler/H2GCN-PyTorch mentioned on GitHubpytorchMPL-2.0Syntology: not harvested report
kdd-submitter/on_local_aggregation mentioned on GitHubpytorchSyntology: not harvested report
sxwee/GNNsIMPL pytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

21 samples harvested: 16 ran (of which 2 constructed an object rather than computing a result; 13 with no instrument failure: 4 honoured, 2 violated, 7 with no contract checked; 3 where Syntology's instrument failed) · 5 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

4ran · honoured contract
2ran · violated contract
2ran · our draft was wrong
1ran; the contract check could not run
7ran
5unverified

Licence: 13 of the 21 samples are pointer only, meaning Syntology does not serve that copy's text. For 7 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 5 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

GemsLab/H2GCN official repository 4 samples, 3 ran (1 instrument-failed) · 1 unverified
Layer in 20 papers h2gcn/models/H2GCN.py ran no licence file found · pointer only · bdae8b092f8e8c19 · report fixture: returned Layer · level recorded 2026-09-22
masked_softmax_cross_entropy in 11 papers h2gcn/models/H2GCN.py ran · honoured contract no licence file found · pointer only · 2fc7117a43f16d67 · report fixture: signature preds:EagerTensor/2/float, labels:EagerTensor/2/float, mask:EagerTensor/1/float · returned EagerTensor · level recorded 2026-09-22
toNumpy h2gcn/models/H2GCN.py ran · our draft was wrong no licence file found · pointer only · d2666218b266548c · report fixture: signature x:SparseTensor/2/float · returned dict · level recorded 2026-09-22
H2GCN h2gcn/models/H2GCN.py unverified no licence file found · pointer only · 8bba7c5bbd4a1809 · report
sxwee/GNNsIMPL community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
H2GCN H2GCN/model/h2gcn.py ran no licence file found · pointer only · 6753abfdba2f2cd6 · report fixture: signature x:Tensor/2/float, adj:Tensor/2/float, adj_2hop:Tensor/2/float · returned Tensor · level recorded 2026-09-24
tkipf/gcn found in paper text by Syntology 9 samples, 6 ran (2 instrument-failed) · 3 unverified
dot in 68 papers gcn/layers.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted MIT (permissive) here; the stored copy is pointer only · bc3a2072a2a9cca3 · report fixture: signature x:EagerTensor/2/float, y:EagerTensor/2/float, sparse:bool/0/bool · returned EagerTensor · level recorded 2026-09-22
get_layer_uid in 30 papers gcn/layers.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · b82968db452628fd · report fixture: signature layer_name:str/0/str · returned int · level recorded 2026-09-22
load_data in 700 papers gcn/utils.py ran · our draft was wrong MIT (permissive) · 36d8bb4858e7bb51 · report fixture: signature dataset_str:str/0/str · returned tuple · level recorded 2026-09-22
masked_accuracy in 14 papers gcn/metrics.py ran · violated contract MIT (permissive) · b89d4728d3c8578d · report fixture: signature preds:EagerTensor/2/float, labels:EagerTensor/2/float, mask:EagerTensor/1/int · returned EagerTensor · level recorded 2026-09-22
masked_softmax_cross_entropy in 11 papers gcn/metrics.py ran · honoured contract MIT (permissive) · d103f29cc58337a2 · report fixture: signature preds:EagerTensor/2/float, labels:EagerTensor/2/float, mask:EagerTensor/1/float · returned EagerTensor · level recorded 2026-09-22
zeros in 26 papers gcn/inits.py ran MIT (permissive) here; the stored copy is pointer only · b190ef9c60b2629a · report fixture: signature shape:tuple/1/int, name:str/0/str · returned ResourceVariable · level recorded 2026-09-24
glorot in 26 papers gcn/inits.py unverified MIT (permissive) here; the stored copy is pointer only · 1b491f3f784b7cb0 · report
sparse_dropout in 30 papers gcn/layers.py unverified MIT (permissive) here; the stored copy is pointer only · 81be8a5f4732fe52 · report
uniform in 58 papers gcn/inits.py unverified MIT (permissive) here; the stored copy is pointer only · b0d1b78daecde7e9 · report
williamleif/graphsage-simple found in paper text by Syntology 1 sample, 1 ran (0 instrument-failed) · 0 unverified
MeanAggregator graphsage/aggregators.py ran no licence file found · pointer only · 3c752dfc33d87577 · report fixture: returned MeanAggregator · level recorded 2026-09-22
PetarV-/GAT found in paper text by Syntology 6 samples, 5 ran (0 instrument-failed) · 1 unverified
adj_to_bias utils/process.py ran MIT (permissive) · 5a0b371866f14f67 · report fixture: signature adj:ndarray/3/float, sizes:ndarray/1/int, nhood:int/0/int · returned ndarray · level recorded 2026-09-25
dfs_split utils/process_ppi.py ran MIT (permissive) · fa7f6abbe630e912 · report fixture: signature adj:csr_matrix/2/int · returned ndarray · level recorded 2026-09-25
parse_index_file in 54 papers utils/process.py ran · honoured contract MIT (permissive) here; the stored copy is pointer only · c1d6392f89c5e495 · report fixture: signature filename:str/0/str · returned list · level recorded, date not recorded
sample_mask in 39 papers
census: 4 implementations from 5 papers, 4 distinct outputs on one shared input
utils/process.py ran · violated contract fingerprinted MIT (permissive) · b32fd748b4bbe3bf · report fixture: signature idx:int/0/int, l:int/0/int · returned ndarray · level recorded 2026-09-22
test in 377 papers utils/process_ppi.py ran MIT (permissive) · cbd09a8688d1836a · report fixture: signature adj:csr_matrix/2/int, mapping:list/1/int · returned bool · level recorded 2026-09-25
find_split utils/process_ppi.py unverified MIT (permissive) · 6beddb1e8d4bd5ff · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. For a generated assembly, an independent peer is a second implementation generated separately from the same paper text; neither implementation read the authors' code. At level 2 the two agreed numerically on the fixture's inputs; level 3 is recorded when property tests derived from the paper also passed on both, and each level 3 card prints how many its report lists. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. A generated assembly's label was set when its code was generated for the method, and no model judged it; its card prints it as "by construction" or "not judged by a model". Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another. Where an assembly's phrase has no constructed figure, a brick that ran is not among the harvested samples this build reads, and whether a brick's run only constructed an object is read from that brick's harvested sample, so the figure was not counted; it does not say there were none.

4 assemblies on this paper: 3 harvested, 1 generated, listed in 2 groups (siblings lifted in one batch form one group).

generated Assembly for H2GCN (graph method: H2GCN), generated from the paper text, not the repository's code. 1 brick; individually on a fixture: 1 ran (1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 3; 4 property tests passed on both).

label matches by construction: the code was generated for this method (not judged by a model)

  1. compute_higher_order_adjacency entry · level 3 · ran · violated contract · dec4a63dcff20bce835472e56044d0a23c34e4d5490e57d13ef6b058c4df577a

3 sibling assemblies for H2GCN (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for H2GCN (graph method: H2GCN). 1 brick; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 3 siblings

  1. H2GCN entry · level 1 · ran · 6753abfdba2f2cd6

harvested Assembly for H2GCN (graph method: H2GCN). 1 brick; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 3 siblings

  1. MeanAggregator entry · level 1 · ran · 3c752dfc33d87577

harvested Assembly for H2GCN (graph method: H2GCN). 4 bricks; individually on a fixture: 3 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified. The assembly has not been run end-to-end.

one of 3 siblings

  1. H2GCN entry · level 0 · unverified · 8bba7c5bbd4a1809
  2. Layer helper · level 1 · ran · bdae8b092f8e8c19
  3. masked_softmax_cross_entropy helper · level 1 · ran · honoured contract · 2fc7117a43f16d67
  4. toNumpy helper · level 1 · ran · our draft was wrong · d2666218b266548c

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2006.11468") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Graph Neural NetworkNode ClassificationNode Classification on Non-Homophilic (Heterophilic) Graphs

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Node Classification Actor H2GCN-2 Accuracy 34.49 ± 1.63 #52 of 62 Archive leaderboard report
Node Classification Actor H2GCN-1 Accuracy 34.31 ± 1.31 #53 of 62 Archive leaderboard report
Node Classification Chameleon H2GCN-2 Accuracy 58.38 ± 1.76 #56 of 61 Archive leaderboard report
Node Classification Chameleon H2GCN-1 Accuracy 52.96 ± 2.09 #57 of 61 Archive leaderboard report
Node Classification Chameleon (60%/20%/20% random splits) H2GCN 1:1 Accuracy 52.30 ± 0.48 #34 of 38 Archive leaderboard report
Node Classification Citeseer (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 77.11 ± 1.57 #13 of 26 Archive leaderboard report
Node Classification Cora (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 87.87 ± 1.20 #15 of 26 Archive leaderboard report
Node Classification Cora (60%/20%/20% random splits) H2GCN 1:1 Accuracy 87.52 ± 0.61 #22 of 33 Archive leaderboard report
Node Classification Cornell H2GCN-2 Accuracy 79.46 ± 4.80 #41 of 60 Archive leaderboard report
Node Classification Cornell H2GCN-1 Accuracy 78.11 ± 6.68 #42 of 60 Archive leaderboard report
Node Classification Cornell (60%/20%/20% random splits) H2GCN 1:1 Accuracy 86.23 ± 4.71 #22 of 36 Archive leaderboard report
Node Classification Penn94 H2GCN Accuracy 81.31 ± 0.60 #21 of 32 Archive leaderboard report
Node Classification PubMed (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 89.49 ± 0.38 #9 of 26 Archive leaderboard report
Node Classification PubMed (60%/20%/20% random splits) H2GCN 1:1 Accuracy 87.78 ± 0.28 #28 of 37 Archive leaderboard report
Node Classification Squirrel H2GCN-2 Accuracy 32.33 ± 1.94 #57 of 59 Archive leaderboard report
Node Classification Squirrel H2GCN-1 Accuracy 28.98 ± 1.97 #59 of 59 Archive leaderboard report
Node Classification Squirrel (60%/20%/20% random splits) H2GCN 1:1 Accuracy 30.39 ± 1.22 #36 of 37 Archive leaderboard report
Node Classification Texas H2GCN-1 Accuracy 83.24 ± 7.07 #42 of 62 Archive leaderboard report
Node Classification Texas H2GCN-2 Accuracy 80.00 ± 6.77 #50 of 62 Archive leaderboard report
Node Classification Texas (60%/20%/20% random splits) H2GCN 1:1 Accuracy 85.90 ± 3.53 #22 of 36 Archive leaderboard report
Node Classification Wisconsin H2GCN-1 Accuracy 84.31 ± 3.70 #45 of 63 Archive leaderboard report
Node Classification Wisconsin H2GCN-2 Accuracy 83.14 ± 4.26 #47 of 63 Archive leaderboard report
Node Classification Wisconsin (60%/20%/20% random splits) H2GCN 1:1 Accuracy 87.5 ± 1.77 #20 of 35 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Chameleon (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 60.11 ± 2.15 #25 of 29 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Chameleon(60%/20%/20% random splits) H2GCN 1:1 Accuracy 52.30 ± 0.48 #28 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Cornell (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 82.70 ± 5.28 #15 of 27 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Cornell (60%/20%/20% random splits) H2GCN 1:1 Accuracy 86.23 ± 4.71 #22 of 33 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Deezer-Europe H2GCN 1:1 Accuracy 67.22±0.90 #5 of 28 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Film(48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 35.70 ± 1.00 #19 of 26 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Penn94 H2GCN 1:1 Accuracy 81.31 ± 0.60 #16 of 28 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Squirrel (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 36.48 ± 1.86 #27 of 29 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Texas(60%/20%/20% random splits) H2GCN 1:1 Accuracy 85.90 ± 3.53 #20 of 32 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Wisconsin (48%/32%/20% fixed splits) H2GCN 1:1 Accuracy 87.65 ± 4.98 #10 of 26 Archive leaderboard report
Node Classification on Non-Homophilic (Heterophilic) Graphs Wisconsin(60%/20%/20% random splits) H2GCN 1:1 Accuracy 87.5 ± 1.77 #20 of 32 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections