{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:KYFTQOSR4XYUTMZK4KYYCXU35N","short_pith_number":"pith:KYFTQOSR","schema_version":"1.0","canonical_sha256":"560b383a51e5f149b32ae2b1815e9beb637729c258012fd27022106187c99e49","source":{"kind":"arxiv","id":"2606.24133","version":1},"attestation_state":"computed","paper":{"title":"Holistic Data Scheduler for LLM Pre-training via Multi-Objective Reinforcement Learning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Chenhao Dang, Jing Ma, Mingjie Liao","submitted_at":"2026-06-23T04:32:46Z","abstract_excerpt":"The composition of training data, governed by the diversity of sources and their mixing strategy, is a cornerstone of Large Language Model (LLM) pre-training. Online Data Mixing (ODM), the technique of adaptively adjusting data mixtures during training, has emerged as a promising direction to improve efficiency. However, existing methods are constrained by their reliance on a singular optimization perspective, which fundamentally overlooks the need for complex LLM pre-training to consider the dynamic data composition from multiple dimensions. To overcome this limitation, we introduce the Holis"},"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":"2606.24133","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-23T04:32:46Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"05acf51fa6bb38afea7713719be43d7ae444d830b5f3b9640d073cb5005d723f","abstract_canon_sha256":"ab5a0c088e27837b423db200df9e7e795d57767a4099ac298b86d453cb9acbe4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-24T01:14:41.935877Z","signature_b64":"1Byf5glXHRPpAbrRjSJQTUZ6aGEazoUYEpvARtnMh1hKWD1wyXE7Gtk28zYFld9T2SVRhAFnKRwTD0TMScb2BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"560b383a51e5f149b32ae2b1815e9beb637729c258012fd27022106187c99e49","last_reissued_at":"2026-06-24T01:14:41.935450Z","signature_status":"signed_v1","first_computed_at":"2026-06-24T01:14:41.935450Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Holistic Data Scheduler for LLM Pre-training via Multi-Objective Reinforcement Learning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Chenhao Dang, Jing Ma, Mingjie Liao","submitted_at":"2026-06-23T04:32:46Z","abstract_excerpt":"The composition of training data, governed by the diversity of sources and their mixing strategy, is a cornerstone of Large Language Model (LLM) pre-training. Online Data Mixing (ODM), the technique of adaptively adjusting data mixtures during training, has emerged as a promising direction to improve efficiency. However, existing methods are constrained by their reliance on a singular optimization perspective, which fundamentally overlooks the need for complex LLM pre-training to consider the dynamic data composition from multiple dimensions. To overcome this limitation, we introduce the Holis"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.24133","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/2606.24133/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":"2606.24133","created_at":"2026-06-24T01:14:41.935515+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.24133v1","created_at":"2026-06-24T01:14:41.935515+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.24133","created_at":"2026-06-24T01:14:41.935515+00:00"},{"alias_kind":"pith_short_12","alias_value":"KYFTQOSR4XYU","created_at":"2026-06-24T01:14:41.935515+00:00"},{"alias_kind":"pith_short_16","alias_value":"KYFTQOSR4XYUTMZK","created_at":"2026-06-24T01:14:41.935515+00:00"},{"alias_kind":"pith_short_8","alias_value":"KYFTQOSR","created_at":"2026-06-24T01:14:41.935515+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/KYFTQOSR4XYUTMZK4KYYCXU35N","json":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N.json","graph_json":"https://pith.science/api/pith-number/KYFTQOSR4XYUTMZK4KYYCXU35N/graph.json","events_json":"https://pith.science/api/pith-number/KYFTQOSR4XYUTMZK4KYYCXU35N/events.json","paper":"https://pith.science/paper/KYFTQOSR"},"agent_actions":{"view_html":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N","download_json":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N.json","view_paper":"https://pith.science/paper/KYFTQOSR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.24133&json=true","fetch_graph":"https://pith.science/api/pith-number/KYFTQOSR4XYUTMZK4KYYCXU35N/graph.json","fetch_events":"https://pith.science/api/pith-number/KYFTQOSR4XYUTMZK4KYYCXU35N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N/action/storage_attestation","attest_author":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N/action/author_attestation","sign_citation":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N/action/citation_signature","submit_replication":"https://pith.science/pith/KYFTQOSR4XYUTMZK4KYYCXU35N/action/replication_record"}},"created_at":"2026-06-24T01:14:41.935515+00:00","updated_at":"2026-06-24T01:14:41.935515+00:00"}