{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JUAI5UFICHS2GOSNKYYMSQZVAK","short_pith_number":"pith:JUAI5UFI","schema_version":"1.0","canonical_sha256":"4d008ed0a811e5a33a4d5630c9433502be06e10179401c641035521b871f3594","source":{"kind":"arxiv","id":"2509.24696","version":2},"attestation_state":"computed","paper":{"title":"T-POP: Test-Time Personalization with Online Preference Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Mingze Kong, Min Zhang, Shuang Qiu, Xiang Li, Yao Shu, Yikun Ban, Zhiwei Shang, Zhiyong Wang, Zhongxiang Dai, Zikun Qu","submitted_at":"2025-09-29T12:28:23Z","abstract_excerpt":"Personalizing large language models (LLMs) to individual user preferences is a critical step beyond generating generically helpful responses. However, current personalization methods are ill-suited for new users, as they typically require either slow, resource-intensive fine-tuning or a substantial amount of pre-existing user data, creating a significant cold-start problem. To address this challenge, we introduce a new paradigm for real-time personalization by learning from online pairwise preference feedback collected during text generation. We propose T-POP (Test-Time Personalization with On"},"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":"2509.24696","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-29T12:28:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7749f782267bcd776033569212541769e099dd5fb8e7d5a564bac77772e0d4c5","abstract_canon_sha256":"4bb0327297814d25d2c9ff51675853ceb9e79279b1db972e926f8febce28df0d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:34.332240Z","signature_b64":"FIKsOgweGFD/Hb6XXmvbgdJ0Wx9bZg28R+k09RLwI9FikIJxOLwqxfGXykVsue4QSzUQUbAEYZc1WN6jrOTfCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4d008ed0a811e5a33a4d5630c9433502be06e10179401c641035521b871f3594","last_reissued_at":"2026-06-02T01:03:34.331706Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:34.331706Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"T-POP: Test-Time Personalization with Online Preference Feedback","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Mingze Kong, Min Zhang, Shuang Qiu, Xiang Li, Yao Shu, Yikun Ban, Zhiwei Shang, Zhiyong Wang, Zhongxiang Dai, Zikun Qu","submitted_at":"2025-09-29T12:28:23Z","abstract_excerpt":"Personalizing large language models (LLMs) to individual user preferences is a critical step beyond generating generically helpful responses. However, current personalization methods are ill-suited for new users, as they typically require either slow, resource-intensive fine-tuning or a substantial amount of pre-existing user data, creating a significant cold-start problem. To address this challenge, we introduce a new paradigm for real-time personalization by learning from online pairwise preference feedback collected during text generation. We propose T-POP (Test-Time Personalization with On"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.24696","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/2509.24696/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":"2509.24696","created_at":"2026-06-02T01:03:34.331773+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.24696v2","created_at":"2026-06-02T01:03:34.331773+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.24696","created_at":"2026-06-02T01:03:34.331773+00:00"},{"alias_kind":"pith_short_12","alias_value":"JUAI5UFICHS2","created_at":"2026-06-02T01:03:34.331773+00:00"},{"alias_kind":"pith_short_16","alias_value":"JUAI5UFICHS2GOSN","created_at":"2026-06-02T01:03:34.331773+00:00"},{"alias_kind":"pith_short_8","alias_value":"JUAI5UFI","created_at":"2026-06-02T01:03:34.331773+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.10991","citing_title":"Test-Time Personalization: A Diagnostic Framework and Probabilistic Fix for Scaling Failures","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2604.17227","citing_title":"Cloud-native and Distributed Systems for Efficient and Scalable Large Language Models -- A Research Agenda","ref_index":151,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK","json":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK.json","graph_json":"https://pith.science/api/pith-number/JUAI5UFICHS2GOSNKYYMSQZVAK/graph.json","events_json":"https://pith.science/api/pith-number/JUAI5UFICHS2GOSNKYYMSQZVAK/events.json","paper":"https://pith.science/paper/JUAI5UFI"},"agent_actions":{"view_html":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK","download_json":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK.json","view_paper":"https://pith.science/paper/JUAI5UFI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.24696&json=true","fetch_graph":"https://pith.science/api/pith-number/JUAI5UFICHS2GOSNKYYMSQZVAK/graph.json","fetch_events":"https://pith.science/api/pith-number/JUAI5UFICHS2GOSNKYYMSQZVAK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK/action/storage_attestation","attest_author":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK/action/author_attestation","sign_citation":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK/action/citation_signature","submit_replication":"https://pith.science/pith/JUAI5UFICHS2GOSNKYYMSQZVAK/action/replication_record"}},"created_at":"2026-06-02T01:03:34.331773+00:00","updated_at":"2026-06-02T01:03:34.331773+00:00"}