{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PDPDJOLOR4INUTOSA76UXJPU5J","short_pith_number":"pith:PDPDJOLO","schema_version":"1.0","canonical_sha256":"78de34b96e8f10da4dd207fd4ba5f4ea6b7cff50edcc8007098581a173a8474b","source":{"kind":"arxiv","id":"2604.06210","version":3},"attestation_state":"computed","paper":{"title":"Distributional Open-Ended Evaluation of LLM Cultural Value Alignment Based on Value Codebook","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"DOVE evaluates LLM cultural alignment by mapping texts to a learned value codebook and comparing distributions with unbalanced optimal transport.","cross_cats":["cs.AI","cs.CY","cs.LG"],"primary_cat":"cs.CL","authors_text":"Hyunjin Hwang, Jaehyeok Lee, Jing Yao, JinYeong Bak, Roy Ka-Wei Lee, Xiaoyuan Yi, Xing Xie","submitted_at":"2026-03-16T08:33:10Z","abstract_excerpt":"As LLMs are globally deployed, aligning their cultural value orientations is critical for safety and user engagement. However, existing benchmarks face the Construct-Composition-Context ($C^3$) challenge: relying on discriminative, multiple-choice formats that probe value knowledge rather than true orientations, overlook subcultural heterogeneity, and mismatch with real-world open-ended generation. We introduce DOVE, a distributional evaluation framework that directly compares human-written text distributions with LLM-generated outputs. DOVE utilizes a rate-distortion variational optimization "},"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":true},"canonical_record":{"source":{"id":"2604.06210","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2026-03-16T08:33:10Z","cross_cats_sorted":["cs.AI","cs.CY","cs.LG"],"title_canon_sha256":"34a68779826653e2793cc7f0f5d9eda708370080df6fb08cad56e2f9428ec0b3","abstract_canon_sha256":"3f1e1e7207143d9fdcdb09dc9d467025029087d94f282a0902c8dc736ad09da2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:52.823584Z","signature_b64":"TJi4kmHcv4BYAtfdzE5yA4+0a52HBBJv0kEjSAgL6y1DAuygRlcYXM1EqaDoanN7pG82ij+dX8Vm+AAB3H9dCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"78de34b96e8f10da4dd207fd4ba5f4ea6b7cff50edcc8007098581a173a8474b","last_reissued_at":"2026-06-02T02:04:52.823079Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:52.823079Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distributional Open-Ended Evaluation of LLM Cultural Value Alignment Based on Value Codebook","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"DOVE evaluates LLM cultural alignment by mapping texts to a learned value codebook and comparing distributions with unbalanced optimal transport.","cross_cats":["cs.AI","cs.CY","cs.LG"],"primary_cat":"cs.CL","authors_text":"Hyunjin Hwang, Jaehyeok Lee, Jing Yao, JinYeong Bak, Roy Ka-Wei Lee, Xiaoyuan Yi, Xing Xie","submitted_at":"2026-03-16T08:33:10Z","abstract_excerpt":"As LLMs are globally deployed, aligning their cultural value orientations is critical for safety and user engagement. However, existing benchmarks face the Construct-Composition-Context ($C^3$) challenge: relying on discriminative, multiple-choice formats that probe value knowledge rather than true orientations, overlook subcultural heterogeneity, and mismatch with real-world open-ended generation. We introduce DOVE, a distributional evaluation framework that directly compares human-written text distributions with LLM-generated outputs. DOVE utilizes a rate-distortion variational optimization "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"DOVE achieves superior predictive validity, attaining a 31.56% correlation with downstream tasks, while maintaining high reliability with as few as 500 samples per culture.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The value codebook constructed from 10K documents via rate-distortion optimization accurately captures true cultural value orientations rather than surface-level semantic patterns, and that distributional differences measured by unbalanced optimal transport reflect genuine alignment rather than generation style artifacts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DOVE constructs a value codebook via rate-distortion variational optimization from 10K documents and measures LLM-human cultural alignment through unbalanced optimal transport, showing 31.56% correlation with downstream tasks and reliability at 500 samples per culture.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"DOVE evaluates LLM cultural alignment by mapping texts to a learned value codebook and comparing distributions with unbalanced optimal transport.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"2de6e300927b5723887d6b61144477ccf8d76556ac2fb571bddcdc163204fc81"},"source":{"id":"2604.06210","kind":"arxiv","version":3},"verdict":{"id":"20fd7bed-436b-4f9a-91da-08b117a11d14","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T10:49:22.828737Z","strongest_claim":"DOVE achieves superior predictive validity, attaining a 31.56% correlation with downstream tasks, while maintaining high reliability with as few as 500 samples per culture.","one_line_summary":"DOVE constructs a value codebook via rate-distortion variational optimization from 10K documents and measures LLM-human cultural alignment through unbalanced optimal transport, showing 31.56% correlation with downstream tasks and reliability at 500 samples per culture.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The value codebook constructed from 10K documents via rate-distortion optimization accurately captures true cultural value orientations rather than surface-level semantic patterns, and that distributional differences measured by unbalanced optimal transport reflect genuine alignment rather than generation style artifacts.","pith_extraction_headline":"DOVE evaluates LLM cultural alignment by mapping texts to a learned value codebook and comparing distributions with unbalanced optimal transport."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.06210/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":2,"snapshot_sha256":"d8e5314fcefacb6f44eab036776b7d75e14a5826aaa64222b2df1f59bcb849d2"},"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":"2604.06210","created_at":"2026-06-02T02:04:52.823132+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.06210v3","created_at":"2026-06-02T02:04:52.823132+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.06210","created_at":"2026-06-02T02:04:52.823132+00:00"},{"alias_kind":"pith_short_12","alias_value":"PDPDJOLOR4IN","created_at":"2026-06-02T02:04:52.823132+00:00"},{"alias_kind":"pith_short_16","alias_value":"PDPDJOLOR4INUTOS","created_at":"2026-06-02T02:04:52.823132+00:00"},{"alias_kind":"pith_short_8","alias_value":"PDPDJOLO","created_at":"2026-06-02T02:04:52.823132+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.22153","citing_title":"When AI Speaks, Whose Values Does It Express? A Cross-Cultural Audit of Individualism-Collectivism Bias in Large Language Models","ref_index":6,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J","json":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J.json","graph_json":"https://pith.science/api/pith-number/PDPDJOLOR4INUTOSA76UXJPU5J/graph.json","events_json":"https://pith.science/api/pith-number/PDPDJOLOR4INUTOSA76UXJPU5J/events.json","paper":"https://pith.science/paper/PDPDJOLO"},"agent_actions":{"view_html":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J","download_json":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J.json","view_paper":"https://pith.science/paper/PDPDJOLO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.06210&json=true","fetch_graph":"https://pith.science/api/pith-number/PDPDJOLOR4INUTOSA76UXJPU5J/graph.json","fetch_events":"https://pith.science/api/pith-number/PDPDJOLOR4INUTOSA76UXJPU5J/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J/action/storage_attestation","attest_author":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J/action/author_attestation","sign_citation":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J/action/citation_signature","submit_replication":"https://pith.science/pith/PDPDJOLOR4INUTOSA76UXJPU5J/action/replication_record"}},"created_at":"2026-06-02T02:04:52.823132+00:00","updated_at":"2026-06-02T02:04:52.823132+00:00"}