{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:AVH2CKNVFMMO3RBZLVBWPASJPT","short_pith_number":"pith:AVH2CKNV","schema_version":"1.0","canonical_sha256":"054fa129b52b18edc4395d436782497ccd21c7b10b06adabba32e452d4d84401","source":{"kind":"arxiv","id":"2402.05133","version":3},"attestation_state":"computed","paper":{"title":"Personalized Language Modeling from Personalized Human Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Liu Leqi, Ruiyang Zhou, Xinyu Li, Zachary C. Lipton","submitted_at":"2024-02-06T04:18:58Z","abstract_excerpt":"Personalized large language models (LLMs) are designed to tailor responses to individual user preferences. While Reinforcement Learning from Human Feedback (RLHF) is a commonly used framework for aligning LLMs with human preferences, vanilla RLHF assumes that all human preferences share the same distribution, preventing fine-tuned LLMs from generating personalized content when user preferences are diverse. In this work, we propose Personalized-RLHF (P-RLHF), an efficient framework that utilizes a lightweight user model to capture individual user preferences and jointly learns the user model an"},"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":"2402.05133","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-06T04:18:58Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"cfa726f7f7db84bc6644ea0f5ab54f1c631b7e13fb09e4891dd6c61499301ca4","abstract_canon_sha256":"c00d6fddc11c469524587fea55746f36e93cc066bcef22d56b32769635658b3b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:45:44.083420Z","signature_b64":"ITsMlvNlPm8ASa8vw7wc89pPlOA7A1MC9wyB0yJEbN3nUu+S2ti4RcHRzFicNIvN1xPjHZ4feL5mZnblnJWQAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"054fa129b52b18edc4395d436782497ccd21c7b10b06adabba32e452d4d84401","last_reissued_at":"2026-07-05T09:45:44.082749Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:45:44.082749Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Personalized Language Modeling from Personalized Human Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Liu Leqi, Ruiyang Zhou, Xinyu Li, Zachary C. Lipton","submitted_at":"2024-02-06T04:18:58Z","abstract_excerpt":"Personalized large language models (LLMs) are designed to tailor responses to individual user preferences. While Reinforcement Learning from Human Feedback (RLHF) is a commonly used framework for aligning LLMs with human preferences, vanilla RLHF assumes that all human preferences share the same distribution, preventing fine-tuned LLMs from generating personalized content when user preferences are diverse. In this work, we propose Personalized-RLHF (P-RLHF), an efficient framework that utilizes a lightweight user model to capture individual user preferences and jointly learns the user model an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.05133","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/2402.05133/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":"2402.05133","created_at":"2026-07-05T09:45:44.082944+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.05133v3","created_at":"2026-07-05T09:45:44.082944+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.05133","created_at":"2026-07-05T09:45:44.082944+00:00"},{"alias_kind":"pith_short_12","alias_value":"AVH2CKNVFMMO","created_at":"2026-07-05T09:45:44.082944+00:00"},{"alias_kind":"pith_short_16","alias_value":"AVH2CKNVFMMO3RBZ","created_at":"2026-07-05T09:45:44.082944+00:00"},{"alias_kind":"pith_short_8","alias_value":"AVH2CKNV","created_at":"2026-07-05T09:45:44.082944+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":16,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17657","citing_title":"Using Cognitive Models to Improve Language Model Simulation of Human Persuasion Games","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01485","citing_title":"CoPersona: Collaborative Persona Graphs for Robust LLM Personalization","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09038","citing_title":"Personalization Meets Safety:Mechanisms,Risks,and Mitigations in Personalized LLMs","ref_index":167,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07988","citing_title":"PAFO: Pareto Fairness Optimization for Personalized Reward Modeling","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02300","citing_title":"Beyond Isolated Behaviors: Hierarchical User Modeling for LLM Personalization","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26072","citing_title":"Active Query Synthesis for Preference Learning","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26969","citing_title":"Recon: Reconstruction-Guided Reasoning Synthesis for User Modeling","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29372","citing_title":"On the Road to Personalized Code Intelligence: Portraiting and Assisting Developers Based on Their In-IDE Behaviors","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00728","citing_title":"From Empathy to Personalized Empathy: Adapting Empathetic Strategies to Individual Users","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2412.08812","citing_title":"Test-Time Alignment via Hypothesis Reweighting","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2510.17881","citing_title":"POPI: Personalizing LLMs via Optimized Natural Language Preference Inference","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11259","citing_title":"Mobile GUI Agent Privacy Personalization with Trajectory Induced Preference Optimization","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09876","citing_title":"Efficient Personalization of Generative User Interfaces","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07629","citing_title":"Behavior Latticing: Inferring User Motivations from Unstructured Interactions","ref_index":54,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07162","citing_title":"CLIPer: Tailoring Diverse User Preference via Classifier-Guided Inference-Time Personalization","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21571","citing_title":"Separable Expert Architecture: Toward Privacy-Preserving LLM Personalization via Composable Adapters and Deletable User Proxies","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT","json":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT.json","graph_json":"https://pith.science/api/pith-number/AVH2CKNVFMMO3RBZLVBWPASJPT/graph.json","events_json":"https://pith.science/api/pith-number/AVH2CKNVFMMO3RBZLVBWPASJPT/events.json","paper":"https://pith.science/paper/AVH2CKNV"},"agent_actions":{"view_html":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT","download_json":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT.json","view_paper":"https://pith.science/paper/AVH2CKNV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.05133&json=true","fetch_graph":"https://pith.science/api/pith-number/AVH2CKNVFMMO3RBZLVBWPASJPT/graph.json","fetch_events":"https://pith.science/api/pith-number/AVH2CKNVFMMO3RBZLVBWPASJPT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT/action/storage_attestation","attest_author":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT/action/author_attestation","sign_citation":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT/action/citation_signature","submit_replication":"https://pith.science/pith/AVH2CKNVFMMO3RBZLVBWPASJPT/action/replication_record"}},"created_at":"2026-07-05T09:45:44.082944+00:00","updated_at":"2026-07-05T09:45:44.082944+00:00"}