{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:EYIY7RNSOHXGQFAFEBQFF7IICW","short_pith_number":"pith:EYIY7RNS","schema_version":"1.0","canonical_sha256":"26118fc5b271ee681405206052fd081589c240b29a8e720abea871069ca2cb42","source":{"kind":"arxiv","id":"2306.03589","version":3},"attestation_state":"computed","paper":{"title":"How does over-squashing affect the power of GNNs?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Andreea Deac, Francesco Di Giovanni, Marc Lackenby, Michael M. Bronstein, Petar Veli\\v{c}kovi\\'c, Siddhartha Mishra, T. Konstantin Rusch","submitted_at":"2023-06-06T11:15:53Z","abstract_excerpt":"Graph Neural Networks (GNNs) are the state-of-the-art model for machine learning on graph-structured data. The most popular class of GNNs operate by exchanging information between adjacent nodes, and are known as Message Passing Neural Networks (MPNNs). Given their widespread use, understanding the expressive power of MPNNs is a key question. However, existing results typically consider settings with uninformative node features. In this paper, we provide a rigorous analysis to determine which function classes of node features can be learned by an MPNN of a given capacity. We do so by measuring"},"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":"2306.03589","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-06-06T11:15:53Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"5a193e10b2320ca2302cc745c6cca0835ec9061c042854a30782105f8c4bf901","abstract_canon_sha256":"9ab3f2150c1748f2ff53e3b32cc9a91ac79577957074638e28a6bb819f5da637"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:44:18.438604Z","signature_b64":"/hB0jPwqSVMS7tzZgZ4OKjb39uZ11DLVfBwm0wUvwa5iRckWEKR8AHM8eRinpwmHjqm6+4/HMkPSlAxcH1MDDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"26118fc5b271ee681405206052fd081589c240b29a8e720abea871069ca2cb42","last_reissued_at":"2026-07-05T07:44:18.437000Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:44:18.437000Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How does over-squashing affect the power of GNNs?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Andreea Deac, Francesco Di Giovanni, Marc Lackenby, Michael M. Bronstein, Petar Veli\\v{c}kovi\\'c, Siddhartha Mishra, T. Konstantin Rusch","submitted_at":"2023-06-06T11:15:53Z","abstract_excerpt":"Graph Neural Networks (GNNs) are the state-of-the-art model for machine learning on graph-structured data. The most popular class of GNNs operate by exchanging information between adjacent nodes, and are known as Message Passing Neural Networks (MPNNs). Given their widespread use, understanding the expressive power of MPNNs is a key question. However, existing results typically consider settings with uninformative node features. In this paper, we provide a rigorous analysis to determine which function classes of node features can be learned by an MPNN of a given capacity. We do so by measuring"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.03589","kind":"arxiv","version":3},"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/2306.03589/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":"2306.03589","created_at":"2026-07-05T07:44:18.438090+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.03589v3","created_at":"2026-07-05T07:44:18.438090+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.03589","created_at":"2026-07-05T07:44:18.438090+00:00"},{"alias_kind":"pith_short_12","alias_value":"EYIY7RNSOHXG","created_at":"2026-07-05T07:44:18.438090+00:00"},{"alias_kind":"pith_short_16","alias_value":"EYIY7RNSOHXGQFAF","created_at":"2026-07-05T07:44:18.438090+00:00"},{"alias_kind":"pith_short_8","alias_value":"EYIY7RNS","created_at":"2026-07-05T07:44:18.438090+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21333","citing_title":"Ramanujan Graph Rewiring with Non Negative Resistance Curvature","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07327","citing_title":"Six Open Questions in Machine-Learned Interatomic Potential Foundation Models","ref_index":122,"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":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW","json":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW.json","graph_json":"https://pith.science/api/pith-number/EYIY7RNSOHXGQFAFEBQFF7IICW/graph.json","events_json":"https://pith.science/api/pith-number/EYIY7RNSOHXGQFAFEBQFF7IICW/events.json","paper":"https://pith.science/paper/EYIY7RNS"},"agent_actions":{"view_html":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW","download_json":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW.json","view_paper":"https://pith.science/paper/EYIY7RNS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.03589&json=true","fetch_graph":"https://pith.science/api/pith-number/EYIY7RNSOHXGQFAFEBQFF7IICW/graph.json","fetch_events":"https://pith.science/api/pith-number/EYIY7RNSOHXGQFAFEBQFF7IICW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW/action/storage_attestation","attest_author":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW/action/author_attestation","sign_citation":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW/action/citation_signature","submit_replication":"https://pith.science/pith/EYIY7RNSOHXGQFAFEBQFF7IICW/action/replication_record"}},"created_at":"2026-07-05T07:44:18.438090+00:00","updated_at":"2026-07-05T07:44:18.438090+00:00"}