{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BL232D7UC4QNZVHH4SGSBLFY2B","short_pith_number":"pith:BL232D7U","schema_version":"1.0","canonical_sha256":"0af5bd0ff41720dcd4e7e48d20acb8d0678d2b759161f806bc18f90355b39b6d","source":{"kind":"arxiv","id":"2512.15765","version":3},"attestation_state":"computed","paper":{"title":"Shapley-based Data Valuation for LLM Alignment via Sequential Preference Optimization","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.GT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Benjamin Heymann, Maxime Vono, M\\'elissa Tamine, Otmane Sakhi, Patrick Loiseau","submitted_at":"2025-12-12T10:13:54Z","abstract_excerpt":"Data valuation is a natural framework for understanding which preference datasets matter most when aligning a Large Language Model (LLM) using multiple sources. The standard game-theoretic approach assigns each dataset a contribution score via the Shapley value. In practice, however, Shapley-based valuation is computationally prohibitive because it requires fine-tuning a separate model for every possible coalition of preference datasets, i.e., an exponential number of alignments. We address this challenge for a broad family of preference-optimization objectives, including DPO and IPO, that lea"},"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":"2512.15765","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-12-12T10:13:54Z","cross_cats_sorted":["cs.GT","stat.ML"],"title_canon_sha256":"e1636424fe004fdb8305b81909f27ed522b635d1262fa69d0009b50bcb5afe90","abstract_canon_sha256":"65c54d1d9780e3cfeb00fff415bef803b8975d401a8494779031dcfe56197205"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T02:17:15.435759Z","signature_b64":"WnIf8aC2MJ8bHxqUTbim3iLt9ysVhvtU++WqiY9alO2iICMtFABP6YCPHhoSZ3O8StIYT3Qcp9BpLeu6NRRTDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0af5bd0ff41720dcd4e7e48d20acb8d0678d2b759161f806bc18f90355b39b6d","last_reissued_at":"2026-07-07T02:17:15.434839Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T02:17:15.434839Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Shapley-based Data Valuation for LLM Alignment via Sequential Preference Optimization","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.GT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Benjamin Heymann, Maxime Vono, M\\'elissa Tamine, Otmane Sakhi, Patrick Loiseau","submitted_at":"2025-12-12T10:13:54Z","abstract_excerpt":"Data valuation is a natural framework for understanding which preference datasets matter most when aligning a Large Language Model (LLM) using multiple sources. The standard game-theoretic approach assigns each dataset a contribution score via the Shapley value. In practice, however, Shapley-based valuation is computationally prohibitive because it requires fine-tuning a separate model for every possible coalition of preference datasets, i.e., an exponential number of alignments. We address this challenge for a broad family of preference-optimization objectives, including DPO and IPO, that lea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.15765","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/2512.15765/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":"2512.15765","created_at":"2026-07-07T02:17:15.434969+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.15765v3","created_at":"2026-07-07T02:17:15.434969+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.15765","created_at":"2026-07-07T02:17:15.434969+00:00"},{"alias_kind":"pith_short_12","alias_value":"BL232D7UC4QN","created_at":"2026-07-07T02:17:15.434969+00:00"},{"alias_kind":"pith_short_16","alias_value":"BL232D7UC4QNZVHH","created_at":"2026-07-07T02:17:15.434969+00:00"},{"alias_kind":"pith_short_8","alias_value":"BL232D7U","created_at":"2026-07-07T02:17:15.434969+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/BL232D7UC4QNZVHH4SGSBLFY2B","json":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B.json","graph_json":"https://pith.science/api/pith-number/BL232D7UC4QNZVHH4SGSBLFY2B/graph.json","events_json":"https://pith.science/api/pith-number/BL232D7UC4QNZVHH4SGSBLFY2B/events.json","paper":"https://pith.science/paper/BL232D7U"},"agent_actions":{"view_html":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B","download_json":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B.json","view_paper":"https://pith.science/paper/BL232D7U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.15765&json=true","fetch_graph":"https://pith.science/api/pith-number/BL232D7UC4QNZVHH4SGSBLFY2B/graph.json","fetch_events":"https://pith.science/api/pith-number/BL232D7UC4QNZVHH4SGSBLFY2B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B/action/storage_attestation","attest_author":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B/action/author_attestation","sign_citation":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B/action/citation_signature","submit_replication":"https://pith.science/pith/BL232D7UC4QNZVHH4SGSBLFY2B/action/replication_record"}},"created_at":"2026-07-07T02:17:15.434969+00:00","updated_at":"2026-07-07T02:17:15.434969+00:00"}