{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PD7OBRIYKZRF6UYCCZ5WWTLMWI","short_pith_number":"pith:PD7OBRIY","schema_version":"1.0","canonical_sha256":"78fee0c51856625f5302167b6b4d6cb2208376573d433e0cb9e5dcb7a78ead80","source":{"kind":"arxiv","id":"2506.04463","version":1},"attestation_state":"computed","paper":{"title":"Aligning Large Language Models with Implicit Preferences from User-Generated Content","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bing Yin, Haodong Wang, Hyokun Yun, Meng Jiang, Ming Zeng, Pei Chen, Priyanka Nigam, Ruijie Wang, Tianyi Liu, Yifan Gao, Zhaoxuan Tan, Zheng Li, Zhihan Zhang","submitted_at":"2025-06-04T21:29:11Z","abstract_excerpt":"Learning from preference feedback is essential for aligning large language models (LLMs) with human values and improving the quality of generated responses. However, existing preference learning methods rely heavily on curated data from humans or advanced LLMs, which is costly and difficult to scale. In this work, we present PUGC, a novel framework that leverages implicit human Preferences in unlabeled User-Generated Content (UGC) to generate preference data. Although UGC is not explicitly created to guide LLMs in generating human-preferred responses, it often reflects valuable insights and im"},"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":"2506.04463","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-04T21:29:11Z","cross_cats_sorted":[],"title_canon_sha256":"d6063f81fc6a3a20039b13ffb5a72ad9325cd9f0fab1d1d3ec8d64e8ffb97f41","abstract_canon_sha256":"f40926208a51af5e436f8eb500486ad47273ec7acb0d32f11cbaf9b0b7db6ebe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:16:14.863702Z","signature_b64":"Ns9aIxWHqhMEG1MMEsci7sNGe0n7FY9XgVrfewyqX0c/DqwtTLCuc+TSQ1EM/NWSwJLKYGtEUgvf2TADqwOTCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"78fee0c51856625f5302167b6b4d6cb2208376573d433e0cb9e5dcb7a78ead80","last_reissued_at":"2026-07-05T11:16:14.863180Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:16:14.863180Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Aligning Large Language Models with Implicit Preferences from User-Generated Content","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bing Yin, Haodong Wang, Hyokun Yun, Meng Jiang, Ming Zeng, Pei Chen, Priyanka Nigam, Ruijie Wang, Tianyi Liu, Yifan Gao, Zhaoxuan Tan, Zheng Li, Zhihan Zhang","submitted_at":"2025-06-04T21:29:11Z","abstract_excerpt":"Learning from preference feedback is essential for aligning large language models (LLMs) with human values and improving the quality of generated responses. However, existing preference learning methods rely heavily on curated data from humans or advanced LLMs, which is costly and difficult to scale. In this work, we present PUGC, a novel framework that leverages implicit human Preferences in unlabeled User-Generated Content (UGC) to generate preference data. Although UGC is not explicitly created to guide LLMs in generating human-preferred responses, it often reflects valuable insights and im"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.04463","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/2506.04463/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":"2506.04463","created_at":"2026-07-05T11:16:14.863254+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.04463v1","created_at":"2026-07-05T11:16:14.863254+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.04463","created_at":"2026-07-05T11:16:14.863254+00:00"},{"alias_kind":"pith_short_12","alias_value":"PD7OBRIYKZRF","created_at":"2026-07-05T11:16:14.863254+00:00"},{"alias_kind":"pith_short_16","alias_value":"PD7OBRIYKZRF6UYC","created_at":"2026-07-05T11:16:14.863254+00:00"},{"alias_kind":"pith_short_8","alias_value":"PD7OBRIY","created_at":"2026-07-05T11:16:14.863254+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2601.21257","citing_title":"MoCo: A One-Stop Shop for Model Collaboration Research","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12479","citing_title":"Meet Dynamic Individual Preferences: Resolving Conflicting Human Value with Paired Fine-Tuning","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI","json":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI.json","graph_json":"https://pith.science/api/pith-number/PD7OBRIYKZRF6UYCCZ5WWTLMWI/graph.json","events_json":"https://pith.science/api/pith-number/PD7OBRIYKZRF6UYCCZ5WWTLMWI/events.json","paper":"https://pith.science/paper/PD7OBRIY"},"agent_actions":{"view_html":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI","download_json":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI.json","view_paper":"https://pith.science/paper/PD7OBRIY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.04463&json=true","fetch_graph":"https://pith.science/api/pith-number/PD7OBRIYKZRF6UYCCZ5WWTLMWI/graph.json","fetch_events":"https://pith.science/api/pith-number/PD7OBRIYKZRF6UYCCZ5WWTLMWI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI/action/storage_attestation","attest_author":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI/action/author_attestation","sign_citation":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI/action/citation_signature","submit_replication":"https://pith.science/pith/PD7OBRIYKZRF6UYCCZ5WWTLMWI/action/replication_record"}},"created_at":"2026-07-05T11:16:14.863254+00:00","updated_at":"2026-07-05T11:16:14.863254+00:00"}