{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:WYJISPVQODPVBKHT7SIGQI6EZV","short_pith_number":"pith:WYJISPVQ","schema_version":"1.0","canonical_sha256":"b612893eb070df50a8f3fc906823c4cd4097569935c83447344e7ce617357e15","source":{"kind":"arxiv","id":"2206.07837","version":4},"attestation_state":"computed","paper":{"title":"Modeling the Data-Generating Process is Necessary for Out-of-Distribution Generalization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Amit Sharma, Emre Kiciman, Jivat Neet Kaur","submitted_at":"2022-06-15T22:35:06Z","abstract_excerpt":"Recent empirical studies on domain generalization (DG) have shown that DG algorithms that perform well on some distribution shifts fail on others, and no state-of-the-art DG algorithm performs consistently well on all shifts. Moreover, real-world data often has multiple distribution shifts over different attributes; hence we introduce multi-attribute distribution shift datasets and find that the accuracy of existing DG algorithms falls even further. To explain these results, we provide a formal characterization of generalization under multi-attribute shifts using a canonical causal graph. Base"},"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":"2206.07837","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-06-15T22:35:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"df17bb3b60db890d1fc875bd97bfc8be797133a8b178679220e0e0c88f3411a1","abstract_canon_sha256":"4a8ab9b2f9687beabd67199460c80a21e45176a7317d5de278a52e7dd27f2949"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:20:20.884163Z","signature_b64":"JEwhwK9zch0Vyuo4iCl2e29jUccEQfKqUCWC4Xr25iYTDY7/DwuuqlquXwlPmZPk+ct7TWHWat/HrLxHXSkiDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b612893eb070df50a8f3fc906823c4cd4097569935c83447344e7ce617357e15","last_reissued_at":"2026-07-05T08:20:20.883729Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:20:20.883729Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Modeling the Data-Generating Process is Necessary for Out-of-Distribution Generalization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Amit Sharma, Emre Kiciman, Jivat Neet Kaur","submitted_at":"2022-06-15T22:35:06Z","abstract_excerpt":"Recent empirical studies on domain generalization (DG) have shown that DG algorithms that perform well on some distribution shifts fail on others, and no state-of-the-art DG algorithm performs consistently well on all shifts. Moreover, real-world data often has multiple distribution shifts over different attributes; hence we introduce multi-attribute distribution shift datasets and find that the accuracy of existing DG algorithms falls even further. To explain these results, we provide a formal characterization of generalization under multi-attribute shifts using a canonical causal graph. Base"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.07837","kind":"arxiv","version":4},"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/2206.07837/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":"2206.07837","created_at":"2026-07-05T08:20:20.883784+00:00"},{"alias_kind":"arxiv_version","alias_value":"2206.07837v4","created_at":"2026-07-05T08:20:20.883784+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.07837","created_at":"2026-07-05T08:20:20.883784+00:00"},{"alias_kind":"pith_short_12","alias_value":"WYJISPVQODPV","created_at":"2026-07-05T08:20:20.883784+00:00"},{"alias_kind":"pith_short_16","alias_value":"WYJISPVQODPVBKHT","created_at":"2026-07-05T08:20:20.883784+00:00"},{"alias_kind":"pith_short_8","alias_value":"WYJISPVQ","created_at":"2026-07-05T08:20:20.883784+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.09433","citing_title":"Mitigating Spurious Correlations in LLMs via Causality-Aware Post-Training","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV","json":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV.json","graph_json":"https://pith.science/api/pith-number/WYJISPVQODPVBKHT7SIGQI6EZV/graph.json","events_json":"https://pith.science/api/pith-number/WYJISPVQODPVBKHT7SIGQI6EZV/events.json","paper":"https://pith.science/paper/WYJISPVQ"},"agent_actions":{"view_html":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV","download_json":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV.json","view_paper":"https://pith.science/paper/WYJISPVQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2206.07837&json=true","fetch_graph":"https://pith.science/api/pith-number/WYJISPVQODPVBKHT7SIGQI6EZV/graph.json","fetch_events":"https://pith.science/api/pith-number/WYJISPVQODPVBKHT7SIGQI6EZV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV/action/storage_attestation","attest_author":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV/action/author_attestation","sign_citation":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV/action/citation_signature","submit_replication":"https://pith.science/pith/WYJISPVQODPVBKHT7SIGQI6EZV/action/replication_record"}},"created_at":"2026-07-05T08:20:20.883784+00:00","updated_at":"2026-07-05T08:20:20.883784+00:00"}