{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:V5T6XRKFLW2PFRBXCIY3WR5NF5","short_pith_number":"pith:V5T6XRKF","schema_version":"1.0","canonical_sha256":"af67ebc5455db4f2c4371231bb47ad2f4f8dda92351f370a8b012c65954bdf40","source":{"kind":"arxiv","id":"2607.20465","version":1},"attestation_state":"computed","paper":{"title":"DataPrep-Bench: Benchmarking LLMs as Training Data Preparators","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bohan Zeng, Conghui He, Hao Liang, Jianzhuo Du, Linzhuang Sun, Meiyi Qiang, Qifeng Cai, Qifeng Xia, Ruichuan An, Sizhe Qiu, Wentao Zhang, Xiaochen Ma, Yibo Lin, Zhaoyang Han","submitted_at":"2026-05-19T02:23:53Z","abstract_excerpt":"The quality of training data fundamentally determines the capabilities of large language models (LLMs), yet no unified benchmark exists to measure how well LLMs, agents, and data-centric workflows actually prepare training data end to end. We view LLM-driven data preparation as comprising two complementary capabilities: data construction, which transforms raw sources into supervised training data, and data quality evaluation, which predicts the training value of candidate datasets before downstream training; throughout, \"quality\" refers to downstream training utility rather than surface-level "},"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":"2607.20465","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T02:23:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"529eac416395574b8fb4db2126248736047bee9daa3285d18356efe9056c31ad","abstract_canon_sha256":"10a03160c833d67bf5f4228c7b258fa657d0d2a8ab9dd7aedd77ecb59a7cfd34"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-24T00:23:17.062151Z","signature_b64":"7G+f48931PWSKTHlYKYPI2rfIRqR+3oZ1/cFNXikrfjE1bTw0+2CcZy+W82uNsA2dkUQ3TDHgqgZ9KTVk3E0Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"af67ebc5455db4f2c4371231bb47ad2f4f8dda92351f370a8b012c65954bdf40","last_reissued_at":"2026-07-24T00:23:17.061257Z","signature_status":"signed_v1","first_computed_at":"2026-07-24T00:23:17.061257Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DataPrep-Bench: Benchmarking LLMs as Training Data Preparators","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bohan Zeng, Conghui He, Hao Liang, Jianzhuo Du, Linzhuang Sun, Meiyi Qiang, Qifeng Cai, Qifeng Xia, Ruichuan An, Sizhe Qiu, Wentao Zhang, Xiaochen Ma, Yibo Lin, Zhaoyang Han","submitted_at":"2026-05-19T02:23:53Z","abstract_excerpt":"The quality of training data fundamentally determines the capabilities of large language models (LLMs), yet no unified benchmark exists to measure how well LLMs, agents, and data-centric workflows actually prepare training data end to end. We view LLM-driven data preparation as comprising two complementary capabilities: data construction, which transforms raw sources into supervised training data, and data quality evaluation, which predicts the training value of candidate datasets before downstream training; throughout, \"quality\" refers to downstream training utility rather than surface-level "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20465","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/2607.20465/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":"2607.20465","created_at":"2026-07-24T00:23:17.061726+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.20465v1","created_at":"2026-07-24T00:23:17.061726+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20465","created_at":"2026-07-24T00:23:17.061726+00:00"},{"alias_kind":"pith_short_12","alias_value":"V5T6XRKFLW2P","created_at":"2026-07-24T00:23:17.061726+00:00"},{"alias_kind":"pith_short_16","alias_value":"V5T6XRKFLW2PFRBX","created_at":"2026-07-24T00:23:17.061726+00:00"},{"alias_kind":"pith_short_8","alias_value":"V5T6XRKF","created_at":"2026-07-24T00:23:17.061726+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/V5T6XRKFLW2PFRBXCIY3WR5NF5","json":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5.json","graph_json":"https://pith.science/api/pith-number/V5T6XRKFLW2PFRBXCIY3WR5NF5/graph.json","events_json":"https://pith.science/api/pith-number/V5T6XRKFLW2PFRBXCIY3WR5NF5/events.json","paper":"https://pith.science/paper/V5T6XRKF"},"agent_actions":{"view_html":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5","download_json":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5.json","view_paper":"https://pith.science/paper/V5T6XRKF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.20465&json=true","fetch_graph":"https://pith.science/api/pith-number/V5T6XRKFLW2PFRBXCIY3WR5NF5/graph.json","fetch_events":"https://pith.science/api/pith-number/V5T6XRKFLW2PFRBXCIY3WR5NF5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5/action/storage_attestation","attest_author":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5/action/author_attestation","sign_citation":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5/action/citation_signature","submit_replication":"https://pith.science/pith/V5T6XRKFLW2PFRBXCIY3WR5NF5/action/replication_record"}},"created_at":"2026-07-24T00:23:17.061726+00:00","updated_at":"2026-07-24T00:23:17.061726+00:00"}