{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/classic-gnns-are-strong-baselines-reassessing","title":"Classic GNNs are Strong Baselines: Reassessing GNNs for Node Classification","arxiv_id":"2406.08993","date":"2024-06-13","proceeding":null,"authors":["Yuankai Luo","Lei Shi","Xiao-Ming Wu"],"abstract":"Graph Transformers (GTs) have recently emerged as popular alternatives to traditional message-passing Graph Neural Networks (GNNs), due to their theoretically superior expressiveness and impressive performance reported on standard node classification benchmarks, often significantly outperforming GNNs. In this paper, we conduct a thorough empirical analysis to reevaluate the performance of three classic GNN models (GCN, GAT, and GraphSAGE) against GTs. Our findings suggest that the previously reported superiority of GTs may have been overstated due to suboptimal hyperparameter configurations in GNNs. Remarkably, with slight hyperparameter tuning, these classic GNN models achieve state-of-the-art performance, matching or even exceeding that of recent GTs across 17 out of the 18 diverse datasets examined. Additionally, we conduct detailed ablation studies to investigate the influence of various GNN configurations, such as normalization, dropout, residual connections, and network depth, on node classification performance. Our study aims to promote a higher standard of empirical rigor in the field of graph machine learning, encouraging more accurate comparisons and evaluations of model capabilities.","url_abs":"https://arxiv.org/abs/2406.08993v2","url_pdf":"https://arxiv.org/pdf/2406.08993v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"classic-gnns-are-strong-baselines-reassessing","repo_url":"https://github.com/LUOyk1999/tunedGNN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"node-classification","task_name":"Node Classification"},{"task_slug":"node-property-prediction","task_name":"Node Property Prediction"}],"methods":[{"method_slug":"gat","method_name":"GAT"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/node-classification-on-amazon-computers-1","task":"Node Classification","dataset":"Amazon Computers","model":"GAT","rank_in_archive_order":1,"of":12,"metrics":{"Accuracy":"94.09±0.37"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-computers-1","task":"Node Classification","dataset":"Amazon Computers","model":"GCN","rank_in_archive_order":2,"of":12,"metrics":{"Accuracy":"93.99±0.12"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-computers-1","task":"Node Classification","dataset":"Amazon Computers","model":"GraphSAGE","rank_in_archive_order":3,"of":12,"metrics":{"Accuracy":"93.25±0.14"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-photo-1","task":"Node Classification","dataset":"Amazon Photo","model":"GraphSAGE","rank_in_archive_order":1,"of":11,"metrics":{"Accuracy":"96.78 ± 0.23"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-photo-1","task":"Node Classification","dataset":"Amazon Photo","model":"GAT","rank_in_archive_order":2,"of":11,"metrics":{"Accuracy":"96.60 ± 0.33"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-photo-1","task":"Node Classification","dataset":"Amazon Photo","model":"GCN","rank_in_archive_order":3,"of":11,"metrics":{"Accuracy":"96.10 ± 0.46"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-citeseer-with-public","task":"Node Classification","dataset":"CiteSeer with Public Split: fixed 20 nodes per class","model":"GCN","rank_in_archive_order":20,"of":40,"metrics":{"Accuracy":"73.14± 0.67"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-coauthor-cs","task":"Node Classification","dataset":"Coauthor CS","model":"GraphSAGE","rank_in_archive_order":3,"of":24,"metrics":{"Accuracy":"96.38±0.11"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-coauthor-physics","task":"Node Classification","dataset":"Coauthor Physics","model":"GCN","rank_in_archive_order":3,"of":14,"metrics":{"Accuracy":"97.46 ± 0.10"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-cora-with-public-split","task":"Node Classification","dataset":"Cora with Public Split: fixed 20 nodes per class","model":"GCN","rank_in_archive_order":6,"of":36,"metrics":{"Accuracy":"85.1 ± 0.7"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-pubmed-with-public","task":"Node Classification","dataset":"PubMed with Public Split: fixed 20 nodes per class","model":"GCN","rank_in_archive_order":8,"of":37,"metrics":{"Accuracy":"81.12 ± 0.52"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-ratings","task":"Node Classification","dataset":"amazon-ratings","model":"GAT","rank_in_archive_order":1,"of":4,"metrics":{"Accuracy (%)":"55.54 ± 0.51"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-ratings","task":"Node Classification","dataset":"amazon-ratings","model":"GraphSAGE","rank_in_archive_order":2,"of":4,"metrics":{"Accuracy (%)":"55.40 ± 0.21"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-amazon-ratings","task":"Node Classification","dataset":"amazon-ratings","model":"GCN","rank_in_archive_order":4,"of":4,"metrics":{"Accuracy (%)":"53.80 ± 0.60"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-minesweeper","task":"Node Classification","dataset":"minesweeper","model":"GCN","rank_in_archive_order":1,"of":4,"metrics":{"AUCROC":"97.86 ± 0.24"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-minesweeper","task":"Node Classification","dataset":"minesweeper","model":"GraphSAGE","rank_in_archive_order":2,"of":4,"metrics":{"AUCROC":"97.77 ± 0.62"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-minesweeper","task":"Node Classification","dataset":"minesweeper","model":"GAT","rank_in_archive_order":3,"of":4,"metrics":{"AUCROC":"97.73 ± 0.73"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-pokec","task":"Node Classification","dataset":"pokec","model":"GCN","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy":"86.33 ± 0.17"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-questions","task":"Node Classification","dataset":"questions","model":"GCN","rank_in_archive_order":2,"of":3,"metrics":{"AUCROC":"79.02 ± 0.60"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-roman-empire","task":"Node Classification","dataset":"roman-empire","model":"GCN","rank_in_archive_order":4,"of":7,"metrics":{"Accuracy (% )":"91.27±0.20"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-arxiv","task":"Node Property Prediction","dataset":"ogbn-arxiv","model":"GCN","rank_in_archive_order":43,"of":86,"metrics":{"Ext. data":"No","Number of params":"1463336","Test Accuracy":"0.7360 ± 0.0018","Validation Accuracy":"0.7447 ± 0.0014"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-arxiv","task":"Node Property Prediction","dataset":"ogbn-arxiv","model":"GraphSAGE","rank_in_archive_order":53,"of":86,"metrics":{"Ext. data":"No","Number of params":"1727272","Test Accuracy":"0.7295 ± 0.0031","Validation Accuracy":"0.7397 ± 0.0015"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-products","task":"Node Property Prediction","dataset":"ogbn-products","model":"GraphSAGE","rank_in_archive_order":26,"of":64,"metrics":{"Ext. data":"No","Number of params":"433047","Test Accuracy":"0.8389 ± 0.0036","Validation Accuracy":"0.9242 ± 0.0029"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-products","task":"Node Property Prediction","dataset":"ogbn-products","model":"GCN","rank_in_archive_order":34,"of":64,"metrics":{"Ext. data":"No","Number of params":"233047","Test Accuracy":"0.8233 ± 0.0019","Validation Accuracy":"0.9224 ± 0.0036"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-proteins","task":"Node Property Prediction","dataset":"ogbn-proteins","model":"GAT","rank_in_archive_order":15,"of":26,"metrics":{"Ext. data":"No","Number of params":"2943472","Test ROC-AUC":"0.8501 ± 0.0046","Validation ROC-AUC":"0.9067 ± 0.0043"},"uses_additional_data":false},{"leaderboard":"/sota/node-property-prediction-on-ogbn-proteins","task":"Node Property Prediction","dataset":"ogbn-proteins","model":"GraphSAGE","rank_in_archive_order":19,"of":26,"metrics":{"Ext. data":"No","Number of params":"2444896","Test ROC-AUC":"0.8221 ± 0.0032","Validation ROC-AUC":"0.8831 ± 0.0044"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2406.08993","atlas_url":"https://app.syntology.ai/?focus=2406.08993","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2406.08993"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/LUOyk1999/tunedGNN","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":2,"unverified":7},"by_repo_kind":{"official":{"samples":9,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"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"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"319c36c17da75fed","entry":"eval_acc","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"large_graph/data_utils.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/large_graph/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"319c36c17da75fed"}},{"code_sha256_prefix":"5b7702a29f6dafaf","entry":"load_fixed_splits","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"large_graph/data_utils.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/large_graph/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5b7702a29f6dafaf"}},{"code_sha256_prefix":"58baf2e0e8e1755c","entry":"class_rand_splits","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"medium_graph/data_utils.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/medium_graph/data_utils.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"58baf2e0e8e1755c"}},{"code_sha256_prefix":"39f7b3a8e45be81b","entry":"eval_f1","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"large_graph/data_utils.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/large_graph/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"39f7b3a8e45be81b"}},{"code_sha256_prefix":"025e94d610e74ffd","entry":"load_fixed_splits","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"medium_graph/data_utils.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/medium_graph/data_utils.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"025e94d610e74ffd"}},{"code_sha256_prefix":"2ca80bf081304f24","entry":"load_hetero_dataset","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"large_graph/dataset.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/large_graph/dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2ca80bf081304f24"}},{"code_sha256_prefix":"54fd4512b8397f45","entry":"load_model","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"large_graph/logger.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/large_graph/logger.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"54fd4512b8397f45"}},{"code_sha256_prefix":"a411b546b0451f0e","entry":"load_wikics_dataset","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"large_graph/dataset.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/large_graph/dataset.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a411b546b0451f0e"}},{"code_sha256_prefix":"f54bada0dd9396ba","entry":"rand_train_test_idx","repo":"LUOyk1999/tunedGNN","repo_kind":"official","path":"medium_graph/data_utils.py","file_url":"https://github.com/LUOyk1999/tunedGNN/blob/HEAD/medium_graph/data_utils.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f54bada0dd9396ba"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}