{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GUMVECVRNQK52AP6AXUULULRV4","short_pith_number":"pith:GUMVECVR","schema_version":"1.0","canonical_sha256":"3519520ab16c15dd01fe05e945d171af09f4b0400be143730f38de2f70dfdfd5","source":{"kind":"arxiv","id":"2601.04885","version":2},"attestation_state":"computed","paper":{"title":"CuMA: Aligning LLMs with Sparse Cultural Values via Demographic-Aware Mixture of Adapters","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ao Sun, Jiachen Zhu, Shu Su, Xiaoyu Wang, Yuheng Jia, Yu Li, Zhe Tan","submitted_at":"2026-01-08T12:30:43Z","abstract_excerpt":"As Large Language Models (LLMs) serve a global audience, alignment must transition from enforcing universal consensus to respecting cultural pluralism. We demonstrate that dense models, when forced to fit conflicting value distributions, suffer from \\textbf{Mean Collapse}, converging to a generic average that fails to represent diverse groups. We attribute this to \\textbf{Cultural Sparsity}, where gradient interference prevents dense parameters from spanning distinct cultural modes. To resolve this, we propose \\textbf{\\textsc{CuMA}} (\\textbf{Cu}ltural \\textbf{M}ixture of \\textbf{A}dapters), a "},"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":"2601.04885","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-08T12:30:43Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"3196e28cec8d189acbb473dff001829a7fd69eb9322a59665c7777141401fc3d","abstract_canon_sha256":"b84f925602bc69fab41f17450e117ccbd93b7a1b3ee6e0983d316062c73078a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-12T01:09:19.937643Z","signature_b64":"b6H273QQyc+ikiuiSAvPghLTUJ80EwwhQVP2fHH1vHoyHy9FduhVAGsnFBzjaq7hLmx9YrMdRmTeLQCdwCYXCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3519520ab16c15dd01fe05e945d171af09f4b0400be143730f38de2f70dfdfd5","last_reissued_at":"2026-06-12T01:09:19.936723Z","signature_status":"signed_v1","first_computed_at":"2026-06-12T01:09:19.936723Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CuMA: Aligning LLMs with Sparse Cultural Values via Demographic-Aware Mixture of Adapters","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ao Sun, Jiachen Zhu, Shu Su, Xiaoyu Wang, Yuheng Jia, Yu Li, Zhe Tan","submitted_at":"2026-01-08T12:30:43Z","abstract_excerpt":"As Large Language Models (LLMs) serve a global audience, alignment must transition from enforcing universal consensus to respecting cultural pluralism. We demonstrate that dense models, when forced to fit conflicting value distributions, suffer from \\textbf{Mean Collapse}, converging to a generic average that fails to represent diverse groups. We attribute this to \\textbf{Cultural Sparsity}, where gradient interference prevents dense parameters from spanning distinct cultural modes. To resolve this, we propose \\textbf{\\textsc{CuMA}} (\\textbf{Cu}ltural \\textbf{M}ixture of \\textbf{A}dapters), a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.04885","kind":"arxiv","version":2},"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/2601.04885/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":"2601.04885","created_at":"2026-06-12T01:09:19.936845+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.04885v2","created_at":"2026-06-12T01:09:19.936845+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.04885","created_at":"2026-06-12T01:09:19.936845+00:00"},{"alias_kind":"pith_short_12","alias_value":"GUMVECVRNQK5","created_at":"2026-06-12T01:09:19.936845+00:00"},{"alias_kind":"pith_short_16","alias_value":"GUMVECVRNQK52AP6","created_at":"2026-06-12T01:09:19.936845+00:00"},{"alias_kind":"pith_short_8","alias_value":"GUMVECVR","created_at":"2026-06-12T01:09:19.936845+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/GUMVECVRNQK52AP6AXUULULRV4","json":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4.json","graph_json":"https://pith.science/api/pith-number/GUMVECVRNQK52AP6AXUULULRV4/graph.json","events_json":"https://pith.science/api/pith-number/GUMVECVRNQK52AP6AXUULULRV4/events.json","paper":"https://pith.science/paper/GUMVECVR"},"agent_actions":{"view_html":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4","download_json":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4.json","view_paper":"https://pith.science/paper/GUMVECVR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.04885&json=true","fetch_graph":"https://pith.science/api/pith-number/GUMVECVRNQK52AP6AXUULULRV4/graph.json","fetch_events":"https://pith.science/api/pith-number/GUMVECVRNQK52AP6AXUULULRV4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4/action/storage_attestation","attest_author":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4/action/author_attestation","sign_citation":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4/action/citation_signature","submit_replication":"https://pith.science/pith/GUMVECVRNQK52AP6AXUULULRV4/action/replication_record"}},"created_at":"2026-06-12T01:09:19.936845+00:00","updated_at":"2026-06-12T01:09:19.936845+00:00"}