{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:RSFE5GEN3TYF64CTDQ7I3UF7ZE","short_pith_number":"pith:RSFE5GEN","schema_version":"1.0","canonical_sha256":"8c8a4e988ddcf05f70531c3e8dd0bfc91d34895e96fb3b90969c8d6e7094acea","source":{"kind":"arxiv","id":"2112.13023","version":1},"attestation_state":"computed","paper":{"title":"DARTS without a Validation Set: Optimizing the Marginal Likelihood","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Binxin Ru, Clare Lyle, Miroslav Fil, Yarin Gal","submitted_at":"2021-12-24T10:16:38Z","abstract_excerpt":"The success of neural architecture search (NAS) has historically been limited by excessive compute requirements. While modern weight-sharing NAS methods such as DARTS are able to finish the search in single-digit GPU days, extracting the final best architecture from the shared weights is notoriously unreliable. Training-Speed-Estimate (TSE), a recently developed generalization estimator with a Bayesian marginal likelihood interpretation, has previously been used in place of the validation loss for gradient-based optimization in DARTS. This prevents the DARTS skip connection collapse, which sig"},"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":"2112.13023","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-12-24T10:16:38Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"322de047ea0b2eb15515c4a21d6c5eba9404c3d0863725486ccfe523e4d626c2","abstract_canon_sha256":"9dfa069630775e9633635f3cd0d4f95c9a9cf7eee99aedf89177fd090509ae73"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:43:49.120642Z","signature_b64":"mX9O4kHko2P/ZSheQuNN1pEBApTgG8wrZP5q8fXannNp1npEDaCRNygfRFxYy6i1DFfpjjDTYsFvFJJ5SSX3BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8c8a4e988ddcf05f70531c3e8dd0bfc91d34895e96fb3b90969c8d6e7094acea","last_reissued_at":"2026-07-05T03:43:49.120161Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:43:49.120161Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DARTS without a Validation Set: Optimizing the Marginal Likelihood","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Binxin Ru, Clare Lyle, Miroslav Fil, Yarin Gal","submitted_at":"2021-12-24T10:16:38Z","abstract_excerpt":"The success of neural architecture search (NAS) has historically been limited by excessive compute requirements. While modern weight-sharing NAS methods such as DARTS are able to finish the search in single-digit GPU days, extracting the final best architecture from the shared weights is notoriously unreliable. Training-Speed-Estimate (TSE), a recently developed generalization estimator with a Bayesian marginal likelihood interpretation, has previously been used in place of the validation loss for gradient-based optimization in DARTS. This prevents the DARTS skip connection collapse, which sig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.13023","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/2112.13023/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":"2112.13023","created_at":"2026-07-05T03:43:49.120219+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.13023v1","created_at":"2026-07-05T03:43:49.120219+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.13023","created_at":"2026-07-05T03:43:49.120219+00:00"},{"alias_kind":"pith_short_12","alias_value":"RSFE5GEN3TYF","created_at":"2026-07-05T03:43:49.120219+00:00"},{"alias_kind":"pith_short_16","alias_value":"RSFE5GEN3TYF64CT","created_at":"2026-07-05T03:43:49.120219+00:00"},{"alias_kind":"pith_short_8","alias_value":"RSFE5GEN","created_at":"2026-07-05T03:43:49.120219+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/RSFE5GEN3TYF64CTDQ7I3UF7ZE","json":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE.json","graph_json":"https://pith.science/api/pith-number/RSFE5GEN3TYF64CTDQ7I3UF7ZE/graph.json","events_json":"https://pith.science/api/pith-number/RSFE5GEN3TYF64CTDQ7I3UF7ZE/events.json","paper":"https://pith.science/paper/RSFE5GEN"},"agent_actions":{"view_html":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE","download_json":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE.json","view_paper":"https://pith.science/paper/RSFE5GEN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.13023&json=true","fetch_graph":"https://pith.science/api/pith-number/RSFE5GEN3TYF64CTDQ7I3UF7ZE/graph.json","fetch_events":"https://pith.science/api/pith-number/RSFE5GEN3TYF64CTDQ7I3UF7ZE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE/action/storage_attestation","attest_author":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE/action/author_attestation","sign_citation":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE/action/citation_signature","submit_replication":"https://pith.science/pith/RSFE5GEN3TYF64CTDQ7I3UF7ZE/action/replication_record"}},"created_at":"2026-07-05T03:43:49.120219+00:00","updated_at":"2026-07-05T03:43:49.120219+00:00"}