Papers › Interpretable Prototype-based Graph Information Bottleneck

Interpretable Prototype-based Graph Information Bottleneck

30 Oct 2023NeurIPS 2023 11arXiv:2310.19906archive 2025-07-28

Sangwoo Seo, Sungwon Kim, Chanyoung Park

The success of Graph Neural Networks (GNNs) has led to a need for understanding their decision-making process and providing explanations for their predictions, which has given rise to explainable AI (XAI) that offers transparent explanations for black-box models. Recently, the use of prototypes has successfully improved the explainability of models by learning prototypes to imply training graphs that affect the prediction. However, these approaches tend to provide prototypes with excessive information from the entire graph, leading to the exclusion of key substructures or the inclusion of irrelevant substructures, which can limit both the interpretability and the performance of the model in downstream tasks. In this work, we propose a novel framework of explainable GNNs, called interpretable Prototype-based Graph Information Bottleneck (PGIB) that incorporates prototype learning within the information bottleneck framework to provide prototypes with the key subgraph from the input graph that is important for the model prediction. This is the first work that incorporates prototype learning into the process of identifying the key subgraphs that have a critical impact on the prediction performance. Extensive experiments, including qualitative analysis, demonstrate that PGIB outperforms state-of-the-art methods in terms of both prediction performance and explainability.

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.

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

Code

Syntology 49 code samples: 48 from 2 repositories linked to this paper, 1 identical to code harvested from another paper's repository. Across all 49: 27 ran (of which 8 constructed an object rather than computing a result; 12 with no instrument failure: 3 honoured, 0 violated, 9 with no contract checked; 15 where Syntology's instrument failed) · 22 unverified. Where Syntology's instrument failed: 12 where our contract draft was wrong; 3 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 47 samples from 1 repository, 25 ran (11 with no instrument failure; 14 where Syntology's instrument failed); community: 1 sample from 1 repository, 1 ran (0 with no instrument failure; 1 where Syntology's instrument failed); 1 identical to code first harvested elsewhere. 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 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 5 of the 27 samples that ran have no recorded date.

sang-woo-seo/pgib officialmentioned in paperpytorchSyntology: 47 samples, 25 ran (14 instrument-failed) · 22 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

49 samples harvested: 27 ran (of which 8 constructed an object rather than computing a result; 12 with no instrument failure: 3 honoured, 0 violated, 9 with no contract checked; 15 where Syntology's instrument failed) · 22 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.

3ran · honoured contract
12ran · our draft was wrong
3ran; the contract check could not run
9ran
22unverified

Licence: 48 of the 49 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 2 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. 1 sample is listed last, in a group without a repository. “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.

sang-woo-seo/pgib official repository 47 samples, 25 ran (14 instrument-failed) · 22 unverified
BA2MotifDataset models/train_gnns.py ran no licence file found · pointer only · b95855d6225bfe22 · report fixture: returned BA2MotifDataset · level recorded 2026-09-22
DataParser models/train_gnns.py ran no licence file found · pointer only · a111427f69c6f21f · report fixture: returned DataParser · level recorded 2026-09-22
GATNet models/train_gnns.py ran no licence file found · pointer only · c30b22e056f3e1c6 · report fixture: returned GATNet · level recorded 2026-09-22
GCNNet models/train_gnns.py ran no licence file found · pointer only · ebb171576db892ad · report fixture: returned GCNNet · level recorded 2026-09-22
GINNet models/train_gnns.py ran no licence file found · pointer only · 8833fd2a02731c27 · report fixture: returned GINNet · level recorded 2026-09-22
GnnBase models/train_gnns.py ran no licence file found · pointer only · 3ee1425cac3fddc2 · report fixture: returned GnnBase · level recorded 2026-09-22
MCTSNode in 6 papers models/train_gnns.py ran no licence file found · pointer only · 0d084866b0972960 · report fixture: returned MCTSNode · level recorded 2026-09-22
SynGraphDataset models/train_gnns.py ran no licence file found · pointer only · bb3abc79825a7c5d · report fixture: returned SynGraphDataset · level recorded 2026-09-22
calculate_distances models/train_gnns.py ran · our draft was wrong fingerprinted no licence file found · pointer only · 547e683509528028 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
calculate_distances proto_join.py ran fingerprinted no licence file found · pointer only · 81b0ee42ca9a58a4 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-24
compute_scores in 20 papers models/train_gnns.py ran · honoured contract no licence file found · pointer only · b98f5b6ae776daf2 · report fixture: signature score_func:callable/0/?, children:list/1/Child · returned list · level recorded 2026-09-22
dataset_bias_split utils.py ran; the contract check could not run (RAISES) no licence file found · pointer only · 6f8818d283b281b4 · report fixture: signature dataset:dict/0/?, args:Args/0/?, bias:float/0/float, split:list/1/int, total:int/0/int · returned tuple · level recorded, date not recorded
evaluate_GC models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · e5576db421883bf9 · report fixture: signature eval_dataloader:list/1/MockBatch, gnnNets:callable/0/?, criterion:callable/0/? · returned dict · level recorded 2026-09-22
get_dataloader in 111 papers models/train_gnns.py ran; the contract check could not run (RAISES) no licence file found · pointer only · 7ff656b3f46349f6 · report fixture: signature dataset:SimpleGraphDataset/0/?, dataset_name:str/0/str, batch_size:int/0/int, random_split_flag:bool/0/bool, data_split_ratio:list/1/float, seed:int/0/int · returned dict · level recorded 2026-09-22
get_readout_layers models/GAT.py ran · our draft was wrong no licence file found · pointer only · fa6332edbe5967b0 · report fixture: signature readout:str/0/str · returned list · level recorded 2026-09-22
gnn_prot_score models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · 7ab4f946cb7fe323 · report fixture: signature coalition:list/1/int, data:callable/0/?, gnnNet:callable/0/?, prototype:Tensor/2/float · returned float · level recorded 2026-09-22
load_SeniGraph models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · 580be1e0f484618e · report fixture: signature dataset_dir:str/0/str, dataset_name:str/0/str · returned SentiGraphDataset · level recorded 2026-09-22
mcts_rollout models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · 911e61201dcf22f0 · report fixture: signature tree_node:MCTSNode/0/?, state_map:dict/0/?, data:callable/0/?, graph:Graph/0/?, score_func:callable/0/? · returned float · level recorded 2026-09-22
print_dataset_info utils.py ran · our draft was wrong no licence file found · pointer only · 441fe701066ef148 · report fixture: signature train_set:list/1/GraphData, val_set:list/1/GraphData, test_set:list/1/GraphData, the:int/0/int · returned dict · level recorded, date not recorded
read_ba2motif_data models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · e5ec0af8c51c8e1d · report fixture: signature folder:str/0/str, prefix:str/0/str · returned list · level recorded 2026-09-22
read_file in 112 papers models/train_gnns.py ran · honoured contract no licence file found · pointer only · 60ccd1c98038811b · report fixture: signature folder:str/0/str, prefix:str/0/str, name:str/0/str · returned ndarray · level recorded 2026-09-22
read_syn_data models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · 86897cb9f3e6a9ac · report fixture: signature folder:str/0/str, prefix:str/0/str · returned Data · level recorded 2026-09-22
split in 67 papers models/train_gnns.py ran; the contract check could not run (RAISES) no licence file found · pointer only · 4af5d2df5e85cf09 · report fixture: signature data:MinimalData/0/?, batch:ndarray/1/int · returned tuple · level recorded 2026-09-22
test_GC models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · 3119dc1d66b9c736 · report fixture: signature test_dataloader:list/1/MinimalBatch, gnnNets:callable/0/?, criterion:callable/0/? · returned tuple · level recorded, date not recorded
undirected_graph models/train_gnns.py ran · our draft was wrong no licence file found · pointer only · dde52a0f318548b4 · report fixture: signature data:MockData/0/? · returned MockData · level recorded 2026-09-22
GnnNets models/train_gnns.py unverified no licence file found · pointer only · f7c887c19d4fed51 · report
MCTSParser models/train_gnns.py unverified no licence file found · pointer only · 9db5d09e971591f5 · report
MUTAGDataset models/train_gnns.py unverified no licence file found · pointer only · 5dc59b60e7651633 · report
ModelParser models/train_gnns.py unverified no licence file found · pointer only · 6a3982a20c6207dd · report
SentiGraphDataset models/train_gnns.py unverified no licence file found · pointer only · be9fd10d144924f1 · report
TUDataset models/train_gnns.py unverified no licence file found · pointer only · eda5a703c89ff5dd · report
TrainParser models/train_gnns.py unverified no licence file found · pointer only · a0b122a25506012a · report
append_record models/train_gnns.py unverified no licence file found · pointer only · bf873f821f332ef2 · report
get_dataset in 285 papers models/train_gnns.py unverified no licence file found · pointer only · 08d682ad45ea4a1a · report
get_model in 368 papers models/train_gnns.py unverified no licence file found · pointer only · f8aac08b803d2945 · report
graph_dataset_generate utils.py unverified no licence file found · pointer only · b5b1840d4cfa103e · report
join_prototypes_by_activations models/train_gnns.py unverified no licence file found · pointer only · df9b7093d8bd9854 · report
joint models/train_gnns.py unverified no licence file found · pointer only · a8624050e2020f05 · report
load_MUTAG models/train_gnns.py unverified no licence file found · pointer only · 81eb3eea84d906e1 · report
load_MolecueNet models/train_gnns.py unverified no licence file found · pointer only · 97dc990b210ff302 · report
load_TUDataset models/train_gnns.py unverified no licence file found · pointer only · 06fadca54eb36919 · report
load_syn_data models/train_gnns.py unverified no licence file found · pointer only · 034edb375a738a45 · report
mcts models/train_gnns.py unverified no licence file found · pointer only · 1dda46bf623e9a0b · report
read_sentigraph_data models/train_gnns.py unverified no licence file found · pointer only · e2b420d6a538dab8 · report
save_best models/train_gnns.py unverified no licence file found · pointer only · 5ae50e21e3acc1a1 · report
train_GC models/train_gnns.py unverified no licence file found · pointer only · 386583600c7eb1b5 · report
warm_only models/train_gnns.py unverified no licence file found · pointer only · fe2aaa5f58bb5787 · report
samyu0304/graph-information-bottleneck-for-subgraph-recognition community 1 sample, 1 ran (1 instrument-failed) · 0 unverified
k_fold in 9 papers graph-classfication/ours_train_eval.py ran · our draft was wrong MIT (permissive) · 8a60d7e1fd8f0dbe · report fixture: signature dataset:MockDataset/0/?, folds:int/0/int · returned tuple · level recorded, date not recorded
Repository not located 1 sample, 1 ran (0 instrument-failed) · 0 unverified
1 sample is identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
num_graphs in 10 papers repository not located ran · honoured contract licence of this copy not recorded · 9da87b0b1605a660 · report fixture: signature data:MockData/0/? · returned int · level recorded, date not recorded

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. 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.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for Interpretable Prototype-based Graph Information Bottleneck (PGIB) (graph method: Interpretable Prototype-based Graph Information Bottleneck (PGIB)). 43 bricks; individually on a fixture: 22 ran (of which 8 constructed an object rather than computing a result; 10 with no instrument failure: 2 honoured, 0 violated, 8 with no contract checked; 12 where Syntology's instrument failed) · 21 unverified. The assembly has not been run end-to-end.

  1. train_GC entry · level 0 · unverified · 386583600c7eb1b5
  2. GnnNets helper · level 0 · unverified · f7c887c19d4fed51
  3. append_record helper · level 0 · unverified · bf873f821f332ef2
  4. evaluate_GC metric · level 1 · ran · our draft was wrong · e5576db421883bf9
  5. get_dataloader helper · level 1 · ran; the contract check could not run (RAISES) · 7ff656b3f46349f6
  6. get_dataset data_transform · level 0 · unverified · 08d682ad45ea4a1a
  7. join_prototypes_by_activations helper · level 0 · unverified · df9b7093d8bd9854
  8. joint update_rule · level 0 · unverified · a8624050e2020f05
  9. mcts helper · level 0 · unverified · 1dda46bf623e9a0b
  10. save_best helper · level 0 · unverified · 5ae50e21e3acc1a1
  11. test_GC metric · level 1 · ran · our draft was wrong · 3119dc1d66b9c736
  12. warm_only update_rule · level 0 · unverified · fe2aaa5f58bb5787
  13. GnnBase helper · level 1 · ran · 3ee1425cac3fddc2
  14. get_model helper · level 0 · unverified · f8aac08b803d2945
  15. load_MUTAG helper · level 0 · unverified · 81eb3eea84d906e1
  16. load_MolecueNet helper · level 0 · unverified · 97dc990b210ff302
  17. load_SeniGraph helper · level 1 · ran · our draft was wrong · 580be1e0f484618e
  18. load_TUDataset helper · level 0 · unverified · 06fadca54eb36919
  19. load_syn_data helper · level 0 · unverified · 034edb375a738a45
  20. calculate_distances helper · level 1 · ran · our draft was wrong · 547e683509528028
  21. MCTSNode helper · level 1 · ran · 0d084866b0972960
  22. gnn_prot_score helper · level 1 · ran · our draft was wrong · 7ab4f946cb7fe323
  23. mcts_rollout helper · level 1 · ran · our draft was wrong · 911e61201dcf22f0
  24. GATNet helper · level 1 · ran · c30b22e056f3e1c6
  25. GCNNet helper · level 1 · ran · ebb171576db892ad
  26. GINNet helper · level 1 · ran · 8833fd2a02731c27
  27. MUTAGDataset helper · level 0 · unverified · 5dc59b60e7651633
  28. SentiGraphDataset helper · level 0 · unverified · be9fd10d144924f1
  29. TUDataset helper · level 0 · unverified · eda5a703c89ff5dd
  30. BA2MotifDataset helper · level 1 · ran · b95855d6225bfe22
  31. SynGraphDataset helper · level 1 · ran · bb3abc79825a7c5d
  32. compute_scores helper · level 1 · ran · honoured contract · b98f5b6ae776daf2
  33. get_readout_layers helper · level 1 · ran · our draft was wrong · fa6332edbe5967b0
  34. read_sentigraph_data helper · level 0 · unverified · e2b420d6a538dab8
  35. undirected_graph helper · level 1 · ran · our draft was wrong · dde52a0f318548b4
  36. read_ba2motif_data helper · level 1 · ran · our draft was wrong · e5ec0af8c51c8e1d
  37. read_syn_data helper · level 1 · ran · our draft was wrong · 86897cb9f3e6a9ac
  38. read_file helper · level 1 · ran · honoured contract · 60ccd1c98038811b
  39. split helper · level 1 · ran; the contract check could not run (RAISES) · 4af5d2df5e85cf09
  40. DataParser helper · level 1 · ran · a111427f69c6f21f
  41. ModelParser helper · level 0 · unverified · 6a3982a20c6207dd
  42. TrainParser helper · level 0 · unverified · a0b122a25506012a
  43. MCTSParser helper · level 0 · unverified · 9db5d09e971591f5

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="2310.19906") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Decision MakingPrediction

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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