{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:A4LR57DYV7TBRHB47PZYWOOJH3","short_pith_number":"pith:A4LR57DY","schema_version":"1.0","canonical_sha256":"07171efc78afe6189c3cfbf38b39c93efea264970b58f4993eb0cfad7b533e77","source":{"kind":"arxiv","id":"2005.01520","version":1},"attestation_state":"computed","paper":{"title":"Demystifying a Dark Art: Understanding Real-World Machine Learning Model Development","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DB"],"primary_cat":"cs.LG","authors_text":"Aditya Parameswaran, Angela Lee, Doris Lee, Doris Xin","submitted_at":"2020-05-04T14:33:39Z","abstract_excerpt":"It is well-known that the process of developing machine learning (ML) workflows is a dark-art; even experts struggle to find an optimal workflow leading to a high accuracy model. Users currently rely on empirical trial-and-error to obtain their own set of battle-tested guidelines to inform their modeling decisions. In this study, we aim to demystify this dark art by understanding how people iterate on ML workflows in practice. We analyze over 475k user-generated workflows on OpenML, an open-source platform for tracking and sharing ML workflows. We find that users often adopt a manual, automate"},"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":"2005.01520","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-05-04T14:33:39Z","cross_cats_sorted":["cs.DB"],"title_canon_sha256":"f2d23e668934a4f8725fcc74c526e5a8e1aed0b987c5989c98d6647860102e87","abstract_canon_sha256":"da7144b4a4cd76241a755cffb0b72008b4004649c064215bc3562d590e31c586"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:00:08.571729Z","signature_b64":"Pj1NM/Suupt+9moLLeO8jrBFzNIXAvwbgjljQSUwZUPU8MP/KT3KcVR4bm06ycpLr5Ox1M5WAW6IG+V7McuUAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"07171efc78afe6189c3cfbf38b39c93efea264970b58f4993eb0cfad7b533e77","last_reissued_at":"2026-07-05T01:00:08.571338Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:00:08.571338Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Demystifying a Dark Art: Understanding Real-World Machine Learning Model Development","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DB"],"primary_cat":"cs.LG","authors_text":"Aditya Parameswaran, Angela Lee, Doris Lee, Doris Xin","submitted_at":"2020-05-04T14:33:39Z","abstract_excerpt":"It is well-known that the process of developing machine learning (ML) workflows is a dark-art; even experts struggle to find an optimal workflow leading to a high accuracy model. Users currently rely on empirical trial-and-error to obtain their own set of battle-tested guidelines to inform their modeling decisions. In this study, we aim to demystify this dark art by understanding how people iterate on ML workflows in practice. We analyze over 475k user-generated workflows on OpenML, an open-source platform for tracking and sharing ML workflows. We find that users often adopt a manual, automate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2005.01520","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/2005.01520/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":"2005.01520","created_at":"2026-07-05T01:00:08.571402+00:00"},{"alias_kind":"arxiv_version","alias_value":"2005.01520v1","created_at":"2026-07-05T01:00:08.571402+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2005.01520","created_at":"2026-07-05T01:00:08.571402+00:00"},{"alias_kind":"pith_short_12","alias_value":"A4LR57DYV7TB","created_at":"2026-07-05T01:00:08.571402+00:00"},{"alias_kind":"pith_short_16","alias_value":"A4LR57DYV7TBRHB4","created_at":"2026-07-05T01:00:08.571402+00:00"},{"alias_kind":"pith_short_8","alias_value":"A4LR57DY","created_at":"2026-07-05T01:00:08.571402+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.01559","citing_title":"How Weight Resampling and Optimizers Shape the Dynamics of Continual Learning and Forgetting in Neural Networks","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3","json":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3.json","graph_json":"https://pith.science/api/pith-number/A4LR57DYV7TBRHB47PZYWOOJH3/graph.json","events_json":"https://pith.science/api/pith-number/A4LR57DYV7TBRHB47PZYWOOJH3/events.json","paper":"https://pith.science/paper/A4LR57DY"},"agent_actions":{"view_html":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3","download_json":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3.json","view_paper":"https://pith.science/paper/A4LR57DY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2005.01520&json=true","fetch_graph":"https://pith.science/api/pith-number/A4LR57DYV7TBRHB47PZYWOOJH3/graph.json","fetch_events":"https://pith.science/api/pith-number/A4LR57DYV7TBRHB47PZYWOOJH3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3/action/storage_attestation","attest_author":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3/action/author_attestation","sign_citation":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3/action/citation_signature","submit_replication":"https://pith.science/pith/A4LR57DYV7TBRHB47PZYWOOJH3/action/replication_record"}},"created_at":"2026-07-05T01:00:08.571402+00:00","updated_at":"2026-07-05T01:00:08.571402+00:00"}