{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:27SWTQYW3G7R3MT6ZQJJ5TTPW4","short_pith_number":"pith:27SWTQYW","schema_version":"1.0","canonical_sha256":"d7e569c316d9bf1db27ecc129ece6fb731ac85a6de9597e5e66c9ac919ed2c60","source":{"kind":"arxiv","id":"2306.04454","version":1},"attestation_state":"computed","paper":{"title":"Training-Free Neural Active Learning with Initialization-Robustness Guarantees","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Apivich Hemachandra, Bryan Kian Hsiang Low, Jasraj Singh, See-kiong Ng, Zhongxiang Dai","submitted_at":"2023-06-07T14:28:42Z","abstract_excerpt":"Existing neural active learning algorithms have aimed to optimize the predictive performance of neural networks (NNs) by selecting data for labelling. However, other than a good predictive performance, being robust against random parameter initializations is also a crucial requirement in safety-critical applications. To this end, we introduce our expected variance with Gaussian processes (EV-GP) criterion for neural active learning, which is theoretically guaranteed to select data points which lead to trained NNs with both (a) good predictive performances and (b) initialization robustness. Imp"},"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":"2306.04454","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-06-07T14:28:42Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"da837278584eb7929fb46575fe1f136dce737c305ff8cbac30ea6acf63d09b61","abstract_canon_sha256":"f267a4d6416e58ea224ab7dc385b02cd64cce3ff7997ec901f87e5a84fc0df21"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:18:30.498532Z","signature_b64":"ifDfggXRGR4URTiBOj+DW8KuPkayPAm5xWPiJF9KJsY09oPvMJ7uDKKUhDi+gtKsTq7HHSQYFRKu38f7HqWsCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7e569c316d9bf1db27ecc129ece6fb731ac85a6de9597e5e66c9ac919ed2c60","last_reissued_at":"2026-07-05T06:18:30.498053Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:18:30.498053Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training-Free Neural Active Learning with Initialization-Robustness Guarantees","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Apivich Hemachandra, Bryan Kian Hsiang Low, Jasraj Singh, See-kiong Ng, Zhongxiang Dai","submitted_at":"2023-06-07T14:28:42Z","abstract_excerpt":"Existing neural active learning algorithms have aimed to optimize the predictive performance of neural networks (NNs) by selecting data for labelling. However, other than a good predictive performance, being robust against random parameter initializations is also a crucial requirement in safety-critical applications. To this end, we introduce our expected variance with Gaussian processes (EV-GP) criterion for neural active learning, which is theoretically guaranteed to select data points which lead to trained NNs with both (a) good predictive performances and (b) initialization robustness. Imp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.04454","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/2306.04454/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":"2306.04454","created_at":"2026-07-05T06:18:30.498110+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.04454v1","created_at":"2026-07-05T06:18:30.498110+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.04454","created_at":"2026-07-05T06:18:30.498110+00:00"},{"alias_kind":"pith_short_12","alias_value":"27SWTQYW3G7R","created_at":"2026-07-05T06:18:30.498110+00:00"},{"alias_kind":"pith_short_16","alias_value":"27SWTQYW3G7R3MT6","created_at":"2026-07-05T06:18:30.498110+00:00"},{"alias_kind":"pith_short_8","alias_value":"27SWTQYW","created_at":"2026-07-05T06:18:30.498110+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.07909","citing_title":"Uncovering Scaling Laws for Large Language Models via Inverse Problems","ref_index":2020,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4","json":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4.json","graph_json":"https://pith.science/api/pith-number/27SWTQYW3G7R3MT6ZQJJ5TTPW4/graph.json","events_json":"https://pith.science/api/pith-number/27SWTQYW3G7R3MT6ZQJJ5TTPW4/events.json","paper":"https://pith.science/paper/27SWTQYW"},"agent_actions":{"view_html":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4","download_json":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4.json","view_paper":"https://pith.science/paper/27SWTQYW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.04454&json=true","fetch_graph":"https://pith.science/api/pith-number/27SWTQYW3G7R3MT6ZQJJ5TTPW4/graph.json","fetch_events":"https://pith.science/api/pith-number/27SWTQYW3G7R3MT6ZQJJ5TTPW4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4/action/storage_attestation","attest_author":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4/action/author_attestation","sign_citation":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4/action/citation_signature","submit_replication":"https://pith.science/pith/27SWTQYW3G7R3MT6ZQJJ5TTPW4/action/replication_record"}},"created_at":"2026-07-05T06:18:30.498110+00:00","updated_at":"2026-07-05T06:18:30.498110+00:00"}