{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:L6CPKTPZINDMZE5QTZIEOBJWOD","short_pith_number":"pith:L6CPKTPZ","schema_version":"1.0","canonical_sha256":"5f84f54df94346cc93b09e5047053670eba2b21cd3f2932374f990f054674e5d","source":{"kind":"arxiv","id":"2606.03866","version":1},"attestation_state":"computed","paper":{"title":"Taiji: Pareto Optimal Policy Optimization with Semantics-IDs Trade-off for Industrial LLM-Enhanced Recommendation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.IR","authors_text":"Chi Lu, Jing Yao, Kun Gai, Peng Jiang, Yuecheng Li, Zeyu Song","submitted_at":"2026-06-02T16:39:06Z","abstract_excerpt":"Scaling recommender systems via large language models (LLMs) has become a prominent trend in the industry. However, aligning the LLM's semantic space with the recommender's ID space via post-training (e.g., SFT and RL) remains challenging. Existing LLM4Rec paradigms are bottlenecked by two main issues: (1) the difficulty of measuring and improving chain-of-thought (CoT) quality in open-domain recommendation during SFT, and (2) the neglect of the trade-off between LLM semantic rewards and recommendation preference rewards during RL alignment. Inspired by these challenges, we present Taiji, a no"},"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":"2606.03866","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2026-06-02T16:39:06Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"0cdb92198bd7b3d25cd8eb4e57ddfa23bf8e2f1d620c1618289c8a1b3f1eddcf","abstract_canon_sha256":"bbf06408964faee0c855c72e664bd26d6a59e4292a9a57ff0850167ee847544d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T02:06:04.738877Z","signature_b64":"8T6Cn3HOTAdy5Z83OQZTIkMyQj1sc5pG+8nGldduXgMe2fpk09Nc38bEnNhtEUUqk9njBNejO8pVvV8wC6MSAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5f84f54df94346cc93b09e5047053670eba2b21cd3f2932374f990f054674e5d","last_reissued_at":"2026-06-03T02:06:04.738489Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T02:06:04.738489Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Taiji: Pareto Optimal Policy Optimization with Semantics-IDs Trade-off for Industrial LLM-Enhanced Recommendation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.IR","authors_text":"Chi Lu, Jing Yao, Kun Gai, Peng Jiang, Yuecheng Li, Zeyu Song","submitted_at":"2026-06-02T16:39:06Z","abstract_excerpt":"Scaling recommender systems via large language models (LLMs) has become a prominent trend in the industry. However, aligning the LLM's semantic space with the recommender's ID space via post-training (e.g., SFT and RL) remains challenging. Existing LLM4Rec paradigms are bottlenecked by two main issues: (1) the difficulty of measuring and improving chain-of-thought (CoT) quality in open-domain recommendation during SFT, and (2) the neglect of the trade-off between LLM semantic rewards and recommendation preference rewards during RL alignment. Inspired by these challenges, we present Taiji, a no"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.03866","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/2606.03866/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":"2606.03866","created_at":"2026-06-03T02:06:04.738544+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.03866v1","created_at":"2026-06-03T02:06:04.738544+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.03866","created_at":"2026-06-03T02:06:04.738544+00:00"},{"alias_kind":"pith_short_12","alias_value":"L6CPKTPZINDM","created_at":"2026-06-03T02:06:04.738544+00:00"},{"alias_kind":"pith_short_16","alias_value":"L6CPKTPZINDMZE5Q","created_at":"2026-06-03T02:06:04.738544+00:00"},{"alias_kind":"pith_short_8","alias_value":"L6CPKTPZ","created_at":"2026-06-03T02:06:04.738544+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/L6CPKTPZINDMZE5QTZIEOBJWOD","json":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD.json","graph_json":"https://pith.science/api/pith-number/L6CPKTPZINDMZE5QTZIEOBJWOD/graph.json","events_json":"https://pith.science/api/pith-number/L6CPKTPZINDMZE5QTZIEOBJWOD/events.json","paper":"https://pith.science/paper/L6CPKTPZ"},"agent_actions":{"view_html":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD","download_json":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD.json","view_paper":"https://pith.science/paper/L6CPKTPZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.03866&json=true","fetch_graph":"https://pith.science/api/pith-number/L6CPKTPZINDMZE5QTZIEOBJWOD/graph.json","fetch_events":"https://pith.science/api/pith-number/L6CPKTPZINDMZE5QTZIEOBJWOD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD/action/storage_attestation","attest_author":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD/action/author_attestation","sign_citation":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD/action/citation_signature","submit_replication":"https://pith.science/pith/L6CPKTPZINDMZE5QTZIEOBJWOD/action/replication_record"}},"created_at":"2026-06-03T02:06:04.738544+00:00","updated_at":"2026-06-03T02:06:04.738544+00:00"}