{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:D7TYCXARY27J4GKM72TIWLMSLU","short_pith_number":"pith:D7TYCXAR","schema_version":"1.0","canonical_sha256":"1fe7815c11c6be9e194cfea68b2d925d0e72b5a3a9f4f590ba53d9656372dd19","source":{"kind":"arxiv","id":"2505.21598","version":1},"attestation_state":"computed","paper":{"title":"Rethinking Data Mixture for Large Language Models: A Comprehensive Survey and New Perspectives","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Congliang Chen, Junchi Yang, Ruoyu Sun, Yajiao Liu","submitted_at":"2025-05-27T16:56:54Z","abstract_excerpt":"Training large language models with data collected from various domains can improve their performance on downstream tasks. However, given a fixed training budget, the sampling proportions of these different domains significantly impact the model's performance. How can we determine the domain weights across different data domains to train the best-performing model within constrained computational resources? In this paper, we provide a comprehensive overview of existing data mixture methods. First, we propose a fine-grained categorization of existing methods, extending beyond the previous offlin"},"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":"2505.21598","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-27T16:56:54Z","cross_cats_sorted":[],"title_canon_sha256":"d2d379c9093de8de6fb104f45396cd8bc04163030bdb5641d34e76f46be8045f","abstract_canon_sha256":"ed41be3cfb55f9e3c3ff431710d7ea4e0bfad9c6a978b552bd0cea4486c5d0c0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:00.807141Z","signature_b64":"mxrD74hNhQCZxzC9lqIkWO8NxjOHem+0ML18H2b0Klyh4uOlPsE0lPzm5UvsmrF+yIBZVcFSOQehhM5SyRV1BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1fe7815c11c6be9e194cfea68b2d925d0e72b5a3a9f4f590ba53d9656372dd19","last_reissued_at":"2026-07-05T11:11:00.806662Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:00.806662Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Data Mixture for Large Language Models: A Comprehensive Survey and New Perspectives","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Congliang Chen, Junchi Yang, Ruoyu Sun, Yajiao Liu","submitted_at":"2025-05-27T16:56:54Z","abstract_excerpt":"Training large language models with data collected from various domains can improve their performance on downstream tasks. However, given a fixed training budget, the sampling proportions of these different domains significantly impact the model's performance. How can we determine the domain weights across different data domains to train the best-performing model within constrained computational resources? In this paper, we provide a comprehensive overview of existing data mixture methods. First, we propose a fine-grained categorization of existing methods, extending beyond the previous offlin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.21598","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/2505.21598/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":"2505.21598","created_at":"2026-07-05T11:11:00.806715+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.21598v1","created_at":"2026-07-05T11:11:00.806715+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.21598","created_at":"2026-07-05T11:11:00.806715+00:00"},{"alias_kind":"pith_short_12","alias_value":"D7TYCXARY27J","created_at":"2026-07-05T11:11:00.806715+00:00"},{"alias_kind":"pith_short_16","alias_value":"D7TYCXARY27J4GKM","created_at":"2026-07-05T11:11:00.806715+00:00"},{"alias_kind":"pith_short_8","alias_value":"D7TYCXAR","created_at":"2026-07-05T11:11:00.806715+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/D7TYCXARY27J4GKM72TIWLMSLU","json":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU.json","graph_json":"https://pith.science/api/pith-number/D7TYCXARY27J4GKM72TIWLMSLU/graph.json","events_json":"https://pith.science/api/pith-number/D7TYCXARY27J4GKM72TIWLMSLU/events.json","paper":"https://pith.science/paper/D7TYCXAR"},"agent_actions":{"view_html":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU","download_json":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU.json","view_paper":"https://pith.science/paper/D7TYCXAR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.21598&json=true","fetch_graph":"https://pith.science/api/pith-number/D7TYCXARY27J4GKM72TIWLMSLU/graph.json","fetch_events":"https://pith.science/api/pith-number/D7TYCXARY27J4GKM72TIWLMSLU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU/action/storage_attestation","attest_author":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU/action/author_attestation","sign_citation":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU/action/citation_signature","submit_replication":"https://pith.science/pith/D7TYCXARY27J4GKM72TIWLMSLU/action/replication_record"}},"created_at":"2026-07-05T11:11:00.806715+00:00","updated_at":"2026-07-05T11:11:00.806715+00:00"}