{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:U37PKMTP3YRYGYOYWJXIJD3XYK","short_pith_number":"pith:U37PKMTP","schema_version":"1.0","canonical_sha256":"a6fef5326fde238361d8b26e848f77c2be8ad650f6fb962893c017d9558fc5b1","source":{"kind":"arxiv","id":"2607.22769","version":1},"attestation_state":"computed","paper":{"title":"DomainPilot: Domain-Level Loss-Guided Two-Stage Data Mixture Optimization for Efficient Language Model Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"He Zhang","submitted_at":"2026-07-24T03:54:58Z","abstract_excerpt":"The training efficacy of large language models (LLMs) is fundamentally constrained by the quality and composition of training data. Existing dynamic data scheduling methods face critical limitations in industrial-scale pretraining and supervised fine-tuning (SFT): data selection incurs prohibitive O(N) costs on terabyte-scale corpora, mixture optimization schemes introduce severe I/O bottlenecks or require training auxiliary reference models, and sample-level reweighting strategies rely on loss signals that conflate noise, difficulty, and novelty.\n  We present DomainPilot, a domain-level loss-"},"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":"2607.22769","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-07-24T03:54:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"60766570892eba5bb490a9e1569100798a3fea5f8bdf8ed63d85e15200f7ea9d","abstract_canon_sha256":"056a4dbc0c63fd648fe77b68342ba0f1c1b18ae15bef87928f05992ec048d57a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-28T00:21:53.440906Z","signature_b64":"ueVvDl+VLXd/5qjUCfDTum6ToFcV2zmNMGa47NAkLZLLeX8+BHr4UFZCJfG1GI/pu8GMX6KbMpfbR6h1giWUCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a6fef5326fde238361d8b26e848f77c2be8ad650f6fb962893c017d9558fc5b1","last_reissued_at":"2026-07-28T00:21:53.440102Z","signature_status":"signed_v1","first_computed_at":"2026-07-28T00:21:53.440102Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DomainPilot: Domain-Level Loss-Guided Two-Stage Data Mixture Optimization for Efficient Language Model Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"He Zhang","submitted_at":"2026-07-24T03:54:58Z","abstract_excerpt":"The training efficacy of large language models (LLMs) is fundamentally constrained by the quality and composition of training data. Existing dynamic data scheduling methods face critical limitations in industrial-scale pretraining and supervised fine-tuning (SFT): data selection incurs prohibitive O(N) costs on terabyte-scale corpora, mixture optimization schemes introduce severe I/O bottlenecks or require training auxiliary reference models, and sample-level reweighting strategies rely on loss signals that conflate noise, difficulty, and novelty.\n  We present DomainPilot, a domain-level loss-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.22769","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/2607.22769/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":"2607.22769","created_at":"2026-07-28T00:21:53.440516+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.22769v1","created_at":"2026-07-28T00:21:53.440516+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.22769","created_at":"2026-07-28T00:21:53.440516+00:00"},{"alias_kind":"pith_short_12","alias_value":"U37PKMTP3YRY","created_at":"2026-07-28T00:21:53.440516+00:00"},{"alias_kind":"pith_short_16","alias_value":"U37PKMTP3YRYGYOY","created_at":"2026-07-28T00:21:53.440516+00:00"},{"alias_kind":"pith_short_8","alias_value":"U37PKMTP","created_at":"2026-07-28T00:21:53.440516+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/U37PKMTP3YRYGYOYWJXIJD3XYK","json":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK.json","graph_json":"https://pith.science/api/pith-number/U37PKMTP3YRYGYOYWJXIJD3XYK/graph.json","events_json":"https://pith.science/api/pith-number/U37PKMTP3YRYGYOYWJXIJD3XYK/events.json","paper":"https://pith.science/paper/U37PKMTP"},"agent_actions":{"view_html":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK","download_json":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK.json","view_paper":"https://pith.science/paper/U37PKMTP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.22769&json=true","fetch_graph":"https://pith.science/api/pith-number/U37PKMTP3YRYGYOYWJXIJD3XYK/graph.json","fetch_events":"https://pith.science/api/pith-number/U37PKMTP3YRYGYOYWJXIJD3XYK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK/action/storage_attestation","attest_author":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK/action/author_attestation","sign_citation":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK/action/citation_signature","submit_replication":"https://pith.science/pith/U37PKMTP3YRYGYOYWJXIJD3XYK/action/replication_record"}},"created_at":"2026-07-28T00:21:53.440516+00:00","updated_at":"2026-07-28T00:21:53.440516+00:00"}