{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:O4EDCQDFFKL6GFTSSY4MB3WGZC","short_pith_number":"pith:O4EDCQDF","schema_version":"1.0","canonical_sha256":"77083140652a97e316729638c0eec6c88ad64914e45facfe8f56f359f8a33b59","source":{"kind":"arxiv","id":"2410.02755","version":3},"attestation_state":"computed","paper":{"title":"GPT-4o as the Gold Standard: A Scalable and General Purpose Approach to Filter Language Model Pretraining Data","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jia Liu, Jifan Zhang, Ness Shroff, Robert Nowak, Ziyue Luo","submitted_at":"2024-10-03T17:58:29Z","abstract_excerpt":"Large language models require vast amounts of high-quality training data, but effective filtering of web-scale datasets remains a significant challenge. This paper demonstrates that GPT-4o is remarkably effective at identifying high-quality training data, but its prohibitive cost makes it impractical at web-scale. We propose SIEVE, a lightweight alternative that matches GPT-4o accuracy at less than 1\\% of the cost. SIEVE can perform up to 500 filtering operations for the cost of one GPT-4o filtering call. The key to SIEVE is a seamless integration of GPT-4o and lightweight text classification "},"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":"2410.02755","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-03T17:58:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1638c4ee87534aa2700ed1bcfab06d3b2a6afd83d9dc7edd1d8a612050ca4f6c","abstract_canon_sha256":"e135486f95de15d62ee27427f8b53120a51e9b6a962d3fa7b2545c084a1ada0a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:07:45.087806Z","signature_b64":"Zv+cO/oyyWJHYj6CiMNmiIinq0Xtl4FbPi+yrVxoOGOVvvnqcAShDbTDgaHoOjBZcxgJrFhPBaQiuPH0aeGnAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"77083140652a97e316729638c0eec6c88ad64914e45facfe8f56f359f8a33b59","last_reissued_at":"2026-07-05T10:07:45.087305Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:07:45.087305Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GPT-4o as the Gold Standard: A Scalable and General Purpose Approach to Filter Language Model Pretraining Data","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jia Liu, Jifan Zhang, Ness Shroff, Robert Nowak, Ziyue Luo","submitted_at":"2024-10-03T17:58:29Z","abstract_excerpt":"Large language models require vast amounts of high-quality training data, but effective filtering of web-scale datasets remains a significant challenge. This paper demonstrates that GPT-4o is remarkably effective at identifying high-quality training data, but its prohibitive cost makes it impractical at web-scale. We propose SIEVE, a lightweight alternative that matches GPT-4o accuracy at less than 1\\% of the cost. SIEVE can perform up to 500 filtering operations for the cost of one GPT-4o filtering call. The key to SIEVE is a seamless integration of GPT-4o and lightweight text classification "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.02755","kind":"arxiv","version":3},"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/2410.02755/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":"2410.02755","created_at":"2026-07-05T10:07:45.087371+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.02755v3","created_at":"2026-07-05T10:07:45.087371+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.02755","created_at":"2026-07-05T10:07:45.087371+00:00"},{"alias_kind":"pith_short_12","alias_value":"O4EDCQDFFKL6","created_at":"2026-07-05T10:07:45.087371+00:00"},{"alias_kind":"pith_short_16","alias_value":"O4EDCQDFFKL6GFTS","created_at":"2026-07-05T10:07:45.087371+00:00"},{"alias_kind":"pith_short_8","alias_value":"O4EDCQDF","created_at":"2026-07-05T10:07:45.087371+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/O4EDCQDFFKL6GFTSSY4MB3WGZC","json":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC.json","graph_json":"https://pith.science/api/pith-number/O4EDCQDFFKL6GFTSSY4MB3WGZC/graph.json","events_json":"https://pith.science/api/pith-number/O4EDCQDFFKL6GFTSSY4MB3WGZC/events.json","paper":"https://pith.science/paper/O4EDCQDF"},"agent_actions":{"view_html":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC","download_json":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC.json","view_paper":"https://pith.science/paper/O4EDCQDF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.02755&json=true","fetch_graph":"https://pith.science/api/pith-number/O4EDCQDFFKL6GFTSSY4MB3WGZC/graph.json","fetch_events":"https://pith.science/api/pith-number/O4EDCQDFFKL6GFTSSY4MB3WGZC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC/action/storage_attestation","attest_author":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC/action/author_attestation","sign_citation":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC/action/citation_signature","submit_replication":"https://pith.science/pith/O4EDCQDFFKL6GFTSSY4MB3WGZC/action/replication_record"}},"created_at":"2026-07-05T10:07:45.087371+00:00","updated_at":"2026-07-05T10:07:45.087371+00:00"}