{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SFG5RHGGENOKXRKGYVHVGBULNT","short_pith_number":"pith:SFG5RHGG","schema_version":"1.0","canonical_sha256":"914dd89cc6235cabc546c54f53068b6ccf687172074af5646701d9f181bbad47","source":{"kind":"arxiv","id":"2504.14194","version":4},"attestation_state":"computed","paper":{"title":"Meta-rater: A Multi-dimensional Data Selection Method for Pre-training Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chi Zhang, Conghui He, Jiahui Peng, Jiantao Qiu, Ren Ma, Tianyi Bai, Xingjian Wei, Xinlin Zhuang, Yinfan Wang, Ying Qian","submitted_at":"2025-04-19T06:12:33Z","abstract_excerpt":"The composition of pre-training datasets for large language models (LLMs) remains largely undisclosed, hindering transparency and efforts to optimize data quality, a critical driver of model performance. Current data selection methods, such as natural language quality assessments, diversity-based filters, and classifier-based approaches, are limited by single-dimensional evaluation or redundancy-focused strategies. To address these gaps, we propose four dimensions to evaluate data quality: professionalism, readability, reasoning, and cleanliness. We further introduce Meta-rater,a multi-dimensi"},"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":"2504.14194","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-04-19T06:12:33Z","cross_cats_sorted":[],"title_canon_sha256":"ec11ce82fcec3c6b1856d8070ef629b7a9dfd1213f52baa15b20b6b9846d4668","abstract_canon_sha256":"1b63ea8256dad22e5d291426c8b5b209cf2f32740f8d9f3c625885d3cdc6fd82"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:49:33.104529Z","signature_b64":"NH/TA2wf0Sd7OUJ0b2d05gbobm9qBFN34oA8kxEeEBPcaktFxNHDK2GqFemOMt4EqhABznHdq3J1UtrZjB7YBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"914dd89cc6235cabc546c54f53068b6ccf687172074af5646701d9f181bbad47","last_reissued_at":"2026-07-05T11:49:33.103971Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:49:33.103971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Meta-rater: A Multi-dimensional Data Selection Method for Pre-training Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chi Zhang, Conghui He, Jiahui Peng, Jiantao Qiu, Ren Ma, Tianyi Bai, Xingjian Wei, Xinlin Zhuang, Yinfan Wang, Ying Qian","submitted_at":"2025-04-19T06:12:33Z","abstract_excerpt":"The composition of pre-training datasets for large language models (LLMs) remains largely undisclosed, hindering transparency and efforts to optimize data quality, a critical driver of model performance. Current data selection methods, such as natural language quality assessments, diversity-based filters, and classifier-based approaches, are limited by single-dimensional evaluation or redundancy-focused strategies. To address these gaps, we propose four dimensions to evaluate data quality: professionalism, readability, reasoning, and cleanliness. We further introduce Meta-rater,a multi-dimensi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.14194","kind":"arxiv","version":4},"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/2504.14194/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":"2504.14194","created_at":"2026-07-05T11:49:33.104053+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.14194v4","created_at":"2026-07-05T11:49:33.104053+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.14194","created_at":"2026-07-05T11:49:33.104053+00:00"},{"alias_kind":"pith_short_12","alias_value":"SFG5RHGGENOK","created_at":"2026-07-05T11:49:33.104053+00:00"},{"alias_kind":"pith_short_16","alias_value":"SFG5RHGGENOKXRKG","created_at":"2026-07-05T11:49:33.104053+00:00"},{"alias_kind":"pith_short_8","alias_value":"SFG5RHGG","created_at":"2026-07-05T11:49:33.104053+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.17650","citing_title":"Measuring Distribution Shift in User Prompts and Its Effects on LLM Performance","ref_index":70,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT","json":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT.json","graph_json":"https://pith.science/api/pith-number/SFG5RHGGENOKXRKGYVHVGBULNT/graph.json","events_json":"https://pith.science/api/pith-number/SFG5RHGGENOKXRKGYVHVGBULNT/events.json","paper":"https://pith.science/paper/SFG5RHGG"},"agent_actions":{"view_html":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT","download_json":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT.json","view_paper":"https://pith.science/paper/SFG5RHGG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.14194&json=true","fetch_graph":"https://pith.science/api/pith-number/SFG5RHGGENOKXRKGYVHVGBULNT/graph.json","fetch_events":"https://pith.science/api/pith-number/SFG5RHGGENOKXRKGYVHVGBULNT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT/action/storage_attestation","attest_author":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT/action/author_attestation","sign_citation":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT/action/citation_signature","submit_replication":"https://pith.science/pith/SFG5RHGGENOKXRKGYVHVGBULNT/action/replication_record"}},"created_at":"2026-07-05T11:49:33.104053+00:00","updated_at":"2026-07-05T11:49:33.104053+00:00"}