{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:AT7GMHSYWUTDAFZPOJW4FLSBQS","short_pith_number":"pith:AT7GMHSY","schema_version":"1.0","canonical_sha256":"04fe661e58b52630172f726dc2ae4184b9588b18de758f6391f1de1f8a85d59b","source":{"kind":"arxiv","id":"1905.10947","version":5},"attestation_state":"computed","paper":{"title":"Graph Neural Networks Exponentially Lose Expressive Power for Node Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Kenta Oono, Taiji Suzuki","submitted_at":"2019-05-27T02:59:06Z","abstract_excerpt":"Graph Neural Networks (graph NNs) are a promising deep learning approach for analyzing graph-structured data. However, it is known that they do not improve (or sometimes worsen) their predictive performance as we pile up many layers and add non-lineality. To tackle this problem, we investigate the expressive power of graph NNs via their asymptotic behaviors as the layer size tends to infinity. Our strategy is to generalize the forward propagation of a Graph Convolutional Network (GCN), which is a popular graph NN variant, as a specific dynamical system. In the case of a GCN, we show that when "},"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":"1905.10947","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-27T02:59:06Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"2d94dde86e707cc717bb4f9d2f47e1f89b43417f8bf8dd098c48cbcbf045c3d4","abstract_canon_sha256":"31868d9bf5eee0ba7395336a65c8cca5ec127d30d6e2ec0a1950090e41bd815e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:04:54.030622Z","signature_b64":"LL29PV/m+gSx8vlU5spKsYCn7uhN+/VoU6+7em1F7sXgHeCPdfsQUwm/r5pfB6+xHyGpubE7CVk80N15YRbvCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"04fe661e58b52630172f726dc2ae4184b9588b18de758f6391f1de1f8a85d59b","last_reissued_at":"2026-07-05T02:04:54.030158Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:04:54.030158Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Graph Neural Networks Exponentially Lose Expressive Power for Node Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Kenta Oono, Taiji Suzuki","submitted_at":"2019-05-27T02:59:06Z","abstract_excerpt":"Graph Neural Networks (graph NNs) are a promising deep learning approach for analyzing graph-structured data. However, it is known that they do not improve (or sometimes worsen) their predictive performance as we pile up many layers and add non-lineality. To tackle this problem, we investigate the expressive power of graph NNs via their asymptotic behaviors as the layer size tends to infinity. Our strategy is to generalize the forward propagation of a Graph Convolutional Network (GCN), which is a popular graph NN variant, as a specific dynamical system. In the case of a GCN, we show that when "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1905.10947","kind":"arxiv","version":5},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1905.10947/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"1905.10947","created_at":"2026-07-05T02:04:54.030235+00:00"},{"alias_kind":"arxiv_version","alias_value":"1905.10947v5","created_at":"2026-07-05T02:04:54.030235+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1905.10947","created_at":"2026-07-05T02:04:54.030235+00:00"},{"alias_kind":"pith_short_12","alias_value":"AT7GMHSYWUTD","created_at":"2026-07-05T02:04:54.030235+00:00"},{"alias_kind":"pith_short_16","alias_value":"AT7GMHSYWUTDAFZP","created_at":"2026-07-05T02:04:54.030235+00:00"},{"alias_kind":"pith_short_8","alias_value":"AT7GMHSY","created_at":"2026-07-05T02:04:54.030235+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22429","citing_title":"Enhancing LLMs for Graph Tasks via Graph-aware LoRA Generation","ref_index":288,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07327","citing_title":"Six Open Questions in Machine-Learned Interatomic Potential Foundation Models","ref_index":118,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02417","citing_title":"Dynamic Spectral Denoising with Global-Context Attention for Multi-Behavior Recommendation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13834","citing_title":"Topology-Preserving Neural Operator Learning via Hodge Decomposition","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24684","citing_title":"Beyond the Aggregation Dilemma: Prior-Retaining Decoupled Learning for Multimodal Graphs","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13834","citing_title":"Topology-Preserving Neural Operator Learning via Hodge Decomposition","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.28161","citing_title":"RopeDreamer: A Kinematic Recurrent State Space Model for Dynamics of Flexible Deformable Linear Objects","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS","json":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS.json","graph_json":"https://pith.science/api/pith-number/AT7GMHSYWUTDAFZPOJW4FLSBQS/graph.json","events_json":"https://pith.science/api/pith-number/AT7GMHSYWUTDAFZPOJW4FLSBQS/events.json","paper":"https://pith.science/paper/AT7GMHSY"},"agent_actions":{"view_html":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS","download_json":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS.json","view_paper":"https://pith.science/paper/AT7GMHSY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1905.10947&json=true","fetch_graph":"https://pith.science/api/pith-number/AT7GMHSYWUTDAFZPOJW4FLSBQS/graph.json","fetch_events":"https://pith.science/api/pith-number/AT7GMHSYWUTDAFZPOJW4FLSBQS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS/action/storage_attestation","attest_author":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS/action/author_attestation","sign_citation":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS/action/citation_signature","submit_replication":"https://pith.science/pith/AT7GMHSYWUTDAFZPOJW4FLSBQS/action/replication_record"}},"created_at":"2026-07-05T02:04:54.030235+00:00","updated_at":"2026-07-05T02:04:54.030235+00:00"}