{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:XMU2RRHNJQIUJCED4T26N4JY4I","short_pith_number":"pith:XMU2RRHN","schema_version":"1.0","canonical_sha256":"bb29a8c4ed4c11448883e4f5e6f138e21f12725470d11e0e1dafd099b4e65825","source":{"kind":"arxiv","id":"1811.05868","version":2},"attestation_state":"computed","paper":{"title":"Pitfalls of Graph Neural Network Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Aleksandar Bojchevski, Maximilian Mumme, Oleksandr Shchur, Stephan G\\\"unnemann","submitted_at":"2018-11-14T15:53:19Z","abstract_excerpt":"Semi-supervised node classification in graphs is a fundamental problem in graph mining, and the recently proposed graph neural networks (GNNs) have achieved unparalleled results on this task. Due to their massive success, GNNs have attracted a lot of attention, and many novel architectures have been put forward. In this paper we show that existing evaluation strategies for GNN models have serious shortcomings. We show that using the same train/validation/test splits of the same datasets, as well as making significant changes to the training procedure (e.g. early stopping criteria) precludes a "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"1811.05868","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-11-14T15:53:19Z","cross_cats_sorted":["cs.SI","stat.ML"],"title_canon_sha256":"95dca9fd6c0748785963c121a35b3898019c66b228954b3d273a43ed7f409b14","abstract_canon_sha256":"ba34a78dfd635877c2d2b2381024190b11ac09f3059a86e774ff9d85b4d82560"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:43:09.331217Z","signature_b64":"J5INwbBKREXh0NNRrHhHEnqif7Tq2onRYRM/TNSxOUNeY4AKgUokHDlNk9FkZN2StP6XHLQn8LxpjfZr6RTrBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bb29a8c4ed4c11448883e4f5e6f138e21f12725470d11e0e1dafd099b4e65825","last_reissued_at":"2026-05-17T23:43:09.330736Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:43:09.330736Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pitfalls of Graph Neural Network Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Aleksandar Bojchevski, Maximilian Mumme, Oleksandr Shchur, Stephan G\\\"unnemann","submitted_at":"2018-11-14T15:53:19Z","abstract_excerpt":"Semi-supervised node classification in graphs is a fundamental problem in graph mining, and the recently proposed graph neural networks (GNNs) have achieved unparalleled results on this task. Due to their massive success, GNNs have attracted a lot of attention, and many novel architectures have been put forward. In this paper we show that existing evaluation strategies for GNN models have serious shortcomings. We show that using the same train/validation/test splits of the same datasets, as well as making significant changes to the training procedure (e.g. early stopping criteria) precludes a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1811.05868","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"1811.05868","created_at":"2026-05-17T23:43:09.330809+00:00"},{"alias_kind":"arxiv_version","alias_value":"1811.05868v2","created_at":"2026-05-17T23:43:09.330809+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1811.05868","created_at":"2026-05-17T23:43:09.330809+00:00"},{"alias_kind":"pith_short_12","alias_value":"XMU2RRHNJQIU","created_at":"2026-05-18T12:33:01.666342+00:00"},{"alias_kind":"pith_short_16","alias_value":"XMU2RRHNJQIUJCED","created_at":"2026-05-18T12:33:01.666342+00:00"},{"alias_kind":"pith_short_8","alias_value":"XMU2RRHN","created_at":"2026-05-18T12:33:01.666342+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":26,"internal_anchor_count":13,"sample":[{"citing_arxiv_id":"2602.11629","citing_title":"GP2F: Cross-Domain Graph Prompting with Adaptive Fusion of Pre-trained Graph Neural Networks","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2405.07406","citing_title":"Machine Unlearning: A Comprehensive Survey","ref_index":151,"is_internal_anchor":true},{"citing_arxiv_id":"2410.15001","citing_title":"FIT-GNN: Faster Inference Time for GNNs that 'FIT' in Memory Using Coarsening","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2506.08618","citing_title":"HSG-12M: A Large-Scale Benchmark of Spatial Multigraphs from the Energy Spectra of Non-Hermitian Crystals","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2603.01388","citing_title":"Invariant-Stratified Propagation for Expressive Graph Neural Networks","ref_index":57,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20248","citing_title":"Graph Transductive Sharpening: Leveraging Unlabeled Predictions in Node Classification","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2605.21247","citing_title":"Graph Navier Stokes Networks","ref_index":62,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17854","citing_title":"Learning over Positive and Negative Edges with Contrastive Message Passing","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2506.08618","citing_title":"HSG-12M: A Large-Scale Benchmark of Spatial Multigraphs from the Energy Spectra of Non-Hermitian Crystals","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2511.06443","citing_title":"How Wide and How Deep? Mitigating Over-Squashing of GNNs via Channel Capacity Constrained Estimation","ref_index":48,"is_internal_anchor":true},{"citing_arxiv_id":"2512.24062","citing_title":"Energy-Balanced Hyperspherical Graph Representation Learning via Structural Binding and Entropic Dispersion","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2602.00407","citing_title":"Fed-Listing: Federated Label Distribution Inference in Graph Neural Networks","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13597","citing_title":"Rethinking Generalization in Graph Neural Networks: A Structural Complexity Perspective","ref_index":33,"is_internal_anchor":true},{"citing_arxiv_id":"2604.02633","citing_title":"Analytic Drift Resister for Non-Exemplar Continual Graph Learning","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11987","citing_title":"Random-Set Graph Neural Networks","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12061","citing_title":"SAGE: A Self-Evolving Agentic Graph-Memory Engine for Structure-Aware Associative Memory","ref_index":108,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27462","citing_title":"Improving Graph Few-shot Learning with Hyperbolic Space and Denoising Diffusion","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09993","citing_title":"Learning Graph Foundation Models on Riemannian Graph-of-Graphs","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09862","citing_title":"UFO: A Unified Flow-Oriented Framework for Robust Continual Graph Learning","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23324","citing_title":"Layer Embedding Deep Fusion Graph Neural Network","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05534","citing_title":"Adversarial Graph Neural Network Benchmarks: Towards Practical and Fair Evaluation","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11473","citing_title":"Learning How Much to Think: Difficulty-Aware Dynamic MoEs for Graph Node Classification","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11257","citing_title":"Unified Graph Prompt Learning via Low-Rank Graph Message Prompting","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08980","citing_title":"Neighbourhood Transformer: Switchable Attention for Monophily-Aware Graph Learning","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06814","citing_title":"From Model to Data (M2D): Shifting Complexity from GNNs to Graphs for Transparent Graph Learning","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I","json":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I.json","graph_json":"https://pith.science/api/pith-number/XMU2RRHNJQIUJCED4T26N4JY4I/graph.json","events_json":"https://pith.science/api/pith-number/XMU2RRHNJQIUJCED4T26N4JY4I/events.json","paper":"https://pith.science/paper/XMU2RRHN"},"agent_actions":{"view_html":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I","download_json":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I.json","view_paper":"https://pith.science/paper/XMU2RRHN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1811.05868&json=true","fetch_graph":"https://pith.science/api/pith-number/XMU2RRHNJQIUJCED4T26N4JY4I/graph.json","fetch_events":"https://pith.science/api/pith-number/XMU2RRHNJQIUJCED4T26N4JY4I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I/action/storage_attestation","attest_author":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I/action/author_attestation","sign_citation":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I/action/citation_signature","submit_replication":"https://pith.science/pith/XMU2RRHNJQIUJCED4T26N4JY4I/action/replication_record"}},"created_at":"2026-05-17T23:43:09.330809+00:00","updated_at":"2026-05-17T23:43:09.330809+00:00"}