{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ISNXR2GCPTNOK352TLBLR2PVMV","short_pith_number":"pith:ISNXR2GC","schema_version":"1.0","canonical_sha256":"449b78e8c27cdae56fba9ac2b8e9f5654f40a346a93a92472d2996a2722a2b74","source":{"kind":"arxiv","id":"2505.14903","version":1},"attestation_state":"computed","paper":{"title":"When to retrain a machine learning model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Coates Mark, Markovich Thomas, Regol Florence, Schwinn Leo, Sprague Kyle","submitted_at":"2025-05-20T20:55:56Z","abstract_excerpt":"A significant challenge in maintaining real-world machine learning models is responding to the continuous and unpredictable evolution of data. Most practitioners are faced with the difficult question: when should I retrain or update my machine learning model? This seemingly straightforward problem is particularly challenging for three reasons: 1) decisions must be made based on very limited information - we usually have access to only a few examples, 2) the nature, extent, and impact of the distribution shift are unknown, and 3) it involves specifying a cost ratio between retraining and poor p"},"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":"2505.14903","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-20T20:55:56Z","cross_cats_sorted":[],"title_canon_sha256":"0c096d969cc84140d9323a43829f6e0b6f1adb110e989d62b6eec1d411856345","abstract_canon_sha256":"7bca7bee18538028c0c99d41610b2019f691fe4eab98cbdba51f8fd7378ce441"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:37.258119Z","signature_b64":"gSfX0d2yQ4X+eSrVh8BAcYYklLozXnV+haVnrdyUWCksh7iyJKUv81P+cddAEK2AGwEDC7RxTYitdkt8w+R9Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"449b78e8c27cdae56fba9ac2b8e9f5654f40a346a93a92472d2996a2722a2b74","last_reissued_at":"2026-07-05T11:06:37.257728Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:37.257728Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When to retrain a machine learning model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Coates Mark, Markovich Thomas, Regol Florence, Schwinn Leo, Sprague Kyle","submitted_at":"2025-05-20T20:55:56Z","abstract_excerpt":"A significant challenge in maintaining real-world machine learning models is responding to the continuous and unpredictable evolution of data. Most practitioners are faced with the difficult question: when should I retrain or update my machine learning model? This seemingly straightforward problem is particularly challenging for three reasons: 1) decisions must be made based on very limited information - we usually have access to only a few examples, 2) the nature, extent, and impact of the distribution shift are unknown, and 3) it involves specifying a cost ratio between retraining and poor p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.14903","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/2505.14903/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":"2505.14903","created_at":"2026-07-05T11:06:37.257784+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.14903v1","created_at":"2026-07-05T11:06:37.257784+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.14903","created_at":"2026-07-05T11:06:37.257784+00:00"},{"alias_kind":"pith_short_12","alias_value":"ISNXR2GCPTNO","created_at":"2026-07-05T11:06:37.257784+00:00"},{"alias_kind":"pith_short_16","alias_value":"ISNXR2GCPTNOK352","created_at":"2026-07-05T11:06:37.257784+00:00"},{"alias_kind":"pith_short_8","alias_value":"ISNXR2GC","created_at":"2026-07-05T11:06:37.257784+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.23954","citing_title":"An empirical evaluation of the risks of AI model updates using clinical data: stability, arbitrariness, and fairness","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07739","citing_title":"Efficient Dataset Selection for Continual Adaptation of Generative Recommenders","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV","json":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV.json","graph_json":"https://pith.science/api/pith-number/ISNXR2GCPTNOK352TLBLR2PVMV/graph.json","events_json":"https://pith.science/api/pith-number/ISNXR2GCPTNOK352TLBLR2PVMV/events.json","paper":"https://pith.science/paper/ISNXR2GC"},"agent_actions":{"view_html":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV","download_json":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV.json","view_paper":"https://pith.science/paper/ISNXR2GC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.14903&json=true","fetch_graph":"https://pith.science/api/pith-number/ISNXR2GCPTNOK352TLBLR2PVMV/graph.json","fetch_events":"https://pith.science/api/pith-number/ISNXR2GCPTNOK352TLBLR2PVMV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV/action/storage_attestation","attest_author":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV/action/author_attestation","sign_citation":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV/action/citation_signature","submit_replication":"https://pith.science/pith/ISNXR2GCPTNOK352TLBLR2PVMV/action/replication_record"}},"created_at":"2026-07-05T11:06:37.257784+00:00","updated_at":"2026-07-05T11:06:37.257784+00:00"}