{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BFSDZE36XNRX5WA3NCLIWELM3D","short_pith_number":"pith:BFSDZE36","schema_version":"1.0","canonical_sha256":"09643c937ebb637ed81b68968b116cd8dc26b079ecf9d768bca464b0f3560156","source":{"kind":"arxiv","id":"2606.22711","version":1},"attestation_state":"computed","paper":{"title":"Beyond Simpson's Paradox: A Cascade of Confounders in AI Agent Pull-Request Co-Authorship","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Haoran Yu, Lifei Liu, Pin Qian, Su Wang, Xiaochong Jiang, Yihang Chen","submitted_at":"2026-06-21T23:16:19Z","abstract_excerpt":"Pooled across five AI coding agents, pull requests (PRs) with a human Co-Authored-By trailer merge less often than purely-autonomous ones (53.8% vs. 79.8%) -- yet this aggregate finding is a textbook Simpson's Paradox. Stratifying 33,596 PRs from the AIDev dataset by agent identity reverses the conclusion: Copilot and Devin show large positive within-agent gaps (+41.2 and +33.5 pp, both p<0.001), while Cursor, Claude Code, and Codex show small effects whose cross-sectional 95% CIs span zero. The paradox is driven entirely by agent composition: Codex, which dominates 64.9% of the dataset, achie"},"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":"2606.22711","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-06-21T23:16:19Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"dd94871027ecc306bc15cc5880ec0f4f6b321741f94da8558cc7ab4d2b64285f","abstract_canon_sha256":"23cad490be6fdc9127fd6911fef983b345cf034bbef0263b4858fea69c155009"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T02:13:45.425123Z","signature_b64":"iuFR6oqL46Cpyav88Ua1oWO6WDwz1VLK8aqyLbh0zTlJ+uP58OMldqXxIP3WN3QekObw5XT7ZyGiCMRQwA4lAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"09643c937ebb637ed81b68968b116cd8dc26b079ecf9d768bca464b0f3560156","last_reissued_at":"2026-06-23T02:13:45.424707Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T02:13:45.424707Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Simpson's Paradox: A Cascade of Confounders in AI Agent Pull-Request Co-Authorship","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Haoran Yu, Lifei Liu, Pin Qian, Su Wang, Xiaochong Jiang, Yihang Chen","submitted_at":"2026-06-21T23:16:19Z","abstract_excerpt":"Pooled across five AI coding agents, pull requests (PRs) with a human Co-Authored-By trailer merge less often than purely-autonomous ones (53.8% vs. 79.8%) -- yet this aggregate finding is a textbook Simpson's Paradox. Stratifying 33,596 PRs from the AIDev dataset by agent identity reverses the conclusion: Copilot and Devin show large positive within-agent gaps (+41.2 and +33.5 pp, both p<0.001), while Cursor, Claude Code, and Codex show small effects whose cross-sectional 95% CIs span zero. The paradox is driven entirely by agent composition: Codex, which dominates 64.9% of the dataset, achie"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22711","kind":"arxiv","version":1},"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/2606.22711/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":"2606.22711","created_at":"2026-06-23T02:13:45.424773+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.22711v1","created_at":"2026-06-23T02:13:45.424773+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22711","created_at":"2026-06-23T02:13:45.424773+00:00"},{"alias_kind":"pith_short_12","alias_value":"BFSDZE36XNRX","created_at":"2026-06-23T02:13:45.424773+00:00"},{"alias_kind":"pith_short_16","alias_value":"BFSDZE36XNRX5WA3","created_at":"2026-06-23T02:13:45.424773+00:00"},{"alias_kind":"pith_short_8","alias_value":"BFSDZE36","created_at":"2026-06-23T02:13:45.424773+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D","json":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D.json","graph_json":"https://pith.science/api/pith-number/BFSDZE36XNRX5WA3NCLIWELM3D/graph.json","events_json":"https://pith.science/api/pith-number/BFSDZE36XNRX5WA3NCLIWELM3D/events.json","paper":"https://pith.science/paper/BFSDZE36"},"agent_actions":{"view_html":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D","download_json":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D.json","view_paper":"https://pith.science/paper/BFSDZE36","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.22711&json=true","fetch_graph":"https://pith.science/api/pith-number/BFSDZE36XNRX5WA3NCLIWELM3D/graph.json","fetch_events":"https://pith.science/api/pith-number/BFSDZE36XNRX5WA3NCLIWELM3D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D/action/storage_attestation","attest_author":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D/action/author_attestation","sign_citation":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D/action/citation_signature","submit_replication":"https://pith.science/pith/BFSDZE36XNRX5WA3NCLIWELM3D/action/replication_record"}},"created_at":"2026-06-23T02:13:45.424773+00:00","updated_at":"2026-06-23T02:13:45.424773+00:00"}