{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FMCDIO2TW27FWJGATDYRQ26ICW","short_pith_number":"pith:FMCDIO2T","schema_version":"1.0","canonical_sha256":"2b04343b53b6be5b24c098f1186bc8158774d82f839439cb475c760876f80394","source":{"kind":"arxiv","id":"2608.08889","version":1},"attestation_state":"computed","paper":{"title":"LLM Reasoning for Subjective Tasks: Failure Modes, Mitigation, and Dynamic Reasoning Routing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Ding Tong, Ishan Gupta, Juncheng Dong, Yuyan Wang","submitted_at":"2026-08-09T20:02:44Z","abstract_excerpt":"Recommendation systems thrive on personalization, where ''correctness'' is rarely a binary truth but a matter of subjective human preference. As Large Language Models (LLMs) are deployed as autonomous verifiers of safety and quality guidelines, they face a distinctive challenge: context-aware preference alignment. Recent gains in Reinforcement Learning with Verifiable Rewards (RLVR) are indexed mostly on objective, mathematical tasks. Through a large-scale study spanning both proprietary and open-source models on four real-world verification tasks from a production recommender platform, we ask"},"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":"2608.08889","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-09T20:02:44Z","cross_cats_sorted":[],"title_canon_sha256":"64386ef5b2c333bd4d762fd5fbfd6c8f72e3a12b29569efd090de52b82917414","abstract_canon_sha256":"27c87cb385ab028ad78d83beb134e6c81baabe10d8245de003dc283a4ab83f7e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:25:38.141610Z","signature_b64":"5MHecb8FF+FfRNNXIzcMSPNKL1FHXiwRZ1Mlkw/IFX/kdhAm49PphUT3Pb/DWVSAMj8QrFbIRosDxzglF3YACA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b04343b53b6be5b24c098f1186bc8158774d82f839439cb475c760876f80394","last_reissued_at":"2026-08-11T01:25:38.138944Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:25:38.138944Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM Reasoning for Subjective Tasks: Failure Modes, Mitigation, and Dynamic Reasoning Routing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Ding Tong, Ishan Gupta, Juncheng Dong, Yuyan Wang","submitted_at":"2026-08-09T20:02:44Z","abstract_excerpt":"Recommendation systems thrive on personalization, where ''correctness'' is rarely a binary truth but a matter of subjective human preference. As Large Language Models (LLMs) are deployed as autonomous verifiers of safety and quality guidelines, they face a distinctive challenge: context-aware preference alignment. Recent gains in Reinforcement Learning with Verifiable Rewards (RLVR) are indexed mostly on objective, mathematical tasks. Through a large-scale study spanning both proprietary and open-source models on four real-world verification tasks from a production recommender platform, we ask"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08889","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/2608.08889/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":"2608.08889","created_at":"2026-08-11T01:25:38.139794+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.08889v1","created_at":"2026-08-11T01:25:38.139794+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08889","created_at":"2026-08-11T01:25:38.139794+00:00"},{"alias_kind":"pith_short_12","alias_value":"FMCDIO2TW27F","created_at":"2026-08-11T01:25:38.139794+00:00"},{"alias_kind":"pith_short_16","alias_value":"FMCDIO2TW27FWJGA","created_at":"2026-08-11T01:25:38.139794+00:00"},{"alias_kind":"pith_short_8","alias_value":"FMCDIO2T","created_at":"2026-08-11T01:25:38.139794+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/FMCDIO2TW27FWJGATDYRQ26ICW","json":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW.json","graph_json":"https://pith.science/api/pith-number/FMCDIO2TW27FWJGATDYRQ26ICW/graph.json","events_json":"https://pith.science/api/pith-number/FMCDIO2TW27FWJGATDYRQ26ICW/events.json","paper":"https://pith.science/paper/FMCDIO2T"},"agent_actions":{"view_html":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW","download_json":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW.json","view_paper":"https://pith.science/paper/FMCDIO2T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.08889&json=true","fetch_graph":"https://pith.science/api/pith-number/FMCDIO2TW27FWJGATDYRQ26ICW/graph.json","fetch_events":"https://pith.science/api/pith-number/FMCDIO2TW27FWJGATDYRQ26ICW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW/action/storage_attestation","attest_author":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW/action/author_attestation","sign_citation":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW/action/citation_signature","submit_replication":"https://pith.science/pith/FMCDIO2TW27FWJGATDYRQ26ICW/action/replication_record"}},"created_at":"2026-08-11T01:25:38.139794+00:00","updated_at":"2026-08-11T01:25:38.139794+00:00"}