{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:QSIFYQJNSZPLFQSS2TVDFF57WF","short_pith_number":"pith:QSIFYQJN","schema_version":"1.0","canonical_sha256":"84905c412d965eb2c252d4ea3297bfb15c4524b43450a610bd018dcfc1e97594","source":{"kind":"arxiv","id":"2109.04173","version":3},"attestation_state":"computed","paper":{"title":"Relating Graph Neural Networks to Structural Causal Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Devendra Singh Dhami, Kristian Kersting, Matej Ze\\v{c}evi\\'c, Petar Veli\\v{c}kovi\\'c","submitted_at":"2021-09-09T11:16:31Z","abstract_excerpt":"Causality can be described in terms of a structural causal model (SCM) that carries information on the variables of interest and their mechanistic relations. For most processes of interest the underlying SCM will only be partially observable, thus causal inference tries leveraging the exposed. Graph neural networks (GNN) as universal approximators on structured input pose a viable candidate for causal learning, suggesting a tighter integration with SCM. To this effect we present a theoretical analysis from first principles that establishes a more general view on neural-causal models, revealing"},"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":"2109.04173","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-09-09T11:16:31Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"c12a640a09bf114c114cd3be1e37994fe84cd8e41b02a8f763126b6df25d6877","abstract_canon_sha256":"d9cfde23091bde9d7281ac836339355ed7429af79205987abc0ba99cb4503580"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:24:58.374467Z","signature_b64":"AyF1ltHqCeznJryco3QPx8bqGW0sqOj7xdBi+Eymm/StlIMxb7B9Py/oIWo3ShFsYer5bt8vKoZhL/h14a/iAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"84905c412d965eb2c252d4ea3297bfb15c4524b43450a610bd018dcfc1e97594","last_reissued_at":"2026-07-05T03:24:58.373982Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:24:58.373982Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Relating Graph Neural Networks to Structural Causal Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Devendra Singh Dhami, Kristian Kersting, Matej Ze\\v{c}evi\\'c, Petar Veli\\v{c}kovi\\'c","submitted_at":"2021-09-09T11:16:31Z","abstract_excerpt":"Causality can be described in terms of a structural causal model (SCM) that carries information on the variables of interest and their mechanistic relations. For most processes of interest the underlying SCM will only be partially observable, thus causal inference tries leveraging the exposed. Graph neural networks (GNN) as universal approximators on structured input pose a viable candidate for causal learning, suggesting a tighter integration with SCM. To this effect we present a theoretical analysis from first principles that establishes a more general view on neural-causal models, revealing"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2109.04173","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/2109.04173/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":"2109.04173","created_at":"2026-07-05T03:24:58.374034+00:00"},{"alias_kind":"arxiv_version","alias_value":"2109.04173v3","created_at":"2026-07-05T03:24:58.374034+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2109.04173","created_at":"2026-07-05T03:24:58.374034+00:00"},{"alias_kind":"pith_short_12","alias_value":"QSIFYQJNSZPL","created_at":"2026-07-05T03:24:58.374034+00:00"},{"alias_kind":"pith_short_16","alias_value":"QSIFYQJNSZPLFQSS","created_at":"2026-07-05T03:24:58.374034+00:00"},{"alias_kind":"pith_short_8","alias_value":"QSIFYQJN","created_at":"2026-07-05T03:24:58.374034+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20526","citing_title":"DeepSWIP: Quotient-WMC Counterfactuals for Neural Probabilistic Logic Programs","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2510.09270","citing_title":"Fast Wasserstein rates for estimating probability distributions of probabilistic graphical models","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2507.13920","citing_title":"Causal Process Models: Reframing Dynamic Causal Graph Discovery as a Reinforcement Learning Problem","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13131","citing_title":"ERPPO: Entropy Regularization-based Proximal Policy Optimization","ref_index":192,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF","json":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF.json","graph_json":"https://pith.science/api/pith-number/QSIFYQJNSZPLFQSS2TVDFF57WF/graph.json","events_json":"https://pith.science/api/pith-number/QSIFYQJNSZPLFQSS2TVDFF57WF/events.json","paper":"https://pith.science/paper/QSIFYQJN"},"agent_actions":{"view_html":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF","download_json":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF.json","view_paper":"https://pith.science/paper/QSIFYQJN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2109.04173&json=true","fetch_graph":"https://pith.science/api/pith-number/QSIFYQJNSZPLFQSS2TVDFF57WF/graph.json","fetch_events":"https://pith.science/api/pith-number/QSIFYQJNSZPLFQSS2TVDFF57WF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF/action/storage_attestation","attest_author":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF/action/author_attestation","sign_citation":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF/action/citation_signature","submit_replication":"https://pith.science/pith/QSIFYQJNSZPLFQSS2TVDFF57WF/action/replication_record"}},"created_at":"2026-07-05T03:24:58.374034+00:00","updated_at":"2026-07-05T03:24:58.374034+00:00"}