{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:7RU7FO3ERSFGWZ3LVDO5ADMVHB","short_pith_number":"pith:7RU7FO3E","schema_version":"1.0","canonical_sha256":"fc69f2bb648c8a6b676ba8ddd00d9538659511032461566d37d0ee2a0e5006cc","source":{"kind":"arxiv","id":"2306.02775","version":3},"attestation_state":"computed","paper":{"title":"Input-gradient space particle inference for neural network ensembles","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Luigi Acerbi, Markus Heinonen, Samuel Kaski, Trung Trinh","submitted_at":"2023-06-05T11:00:11Z","abstract_excerpt":"Deep Ensembles (DEs) demonstrate improved accuracy, calibration and robustness to perturbations over single neural networks partly due to their functional diversity. Particle-based variational inference (ParVI) methods enhance diversity by formalizing a repulsion term based on a network similarity kernel. However, weight-space repulsion is inefficient due to over-parameterization, while direct function-space repulsion has been found to produce little improvement over DEs. To sidestep these difficulties, we propose First-order Repulsive Deep Ensemble (FoRDE), an ensemble learning method based o"},"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":"2306.02775","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-06-05T11:00:11Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"660c900da84481e95657ce3b78238a4f81417b0ea232e90957b4b81d48f26651","abstract_canon_sha256":"b93d717dce1a86faa8e233c2edae1b9bb2513513c5cb24767f7dbcb537315d6f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:52:10.206312Z","signature_b64":"7iRvSsUjhrKdD7646pWAyp2aMYrpCHqx/X4QtRGqk3BsGlD8118BJQWvdkZTOUoqT++2/h1mJ2Jgqz5WTUj6Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc69f2bb648c8a6b676ba8ddd00d9538659511032461566d37d0ee2a0e5006cc","last_reissued_at":"2026-07-05T07:52:10.205851Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:52:10.205851Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Input-gradient space particle inference for neural network ensembles","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Luigi Acerbi, Markus Heinonen, Samuel Kaski, Trung Trinh","submitted_at":"2023-06-05T11:00:11Z","abstract_excerpt":"Deep Ensembles (DEs) demonstrate improved accuracy, calibration and robustness to perturbations over single neural networks partly due to their functional diversity. Particle-based variational inference (ParVI) methods enhance diversity by formalizing a repulsion term based on a network similarity kernel. However, weight-space repulsion is inefficient due to over-parameterization, while direct function-space repulsion has been found to produce little improvement over DEs. To sidestep these difficulties, we propose First-order Repulsive Deep Ensemble (FoRDE), an ensemble learning method based o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.02775","kind":"arxiv","version":3},"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/2306.02775/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":"2306.02775","created_at":"2026-07-05T07:52:10.205926+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.02775v3","created_at":"2026-07-05T07:52:10.205926+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.02775","created_at":"2026-07-05T07:52:10.205926+00:00"},{"alias_kind":"pith_short_12","alias_value":"7RU7FO3ERSFG","created_at":"2026-07-05T07:52:10.205926+00:00"},{"alias_kind":"pith_short_16","alias_value":"7RU7FO3ERSFGWZ3L","created_at":"2026-07-05T07:52:10.205926+00:00"},{"alias_kind":"pith_short_8","alias_value":"7RU7FO3E","created_at":"2026-07-05T07:52:10.205926+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.01334","citing_title":"Symbolic or Numerical? Understanding Physics Problem Solving in Reasoning LLMs","ref_index":13,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB","json":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB.json","graph_json":"https://pith.science/api/pith-number/7RU7FO3ERSFGWZ3LVDO5ADMVHB/graph.json","events_json":"https://pith.science/api/pith-number/7RU7FO3ERSFGWZ3LVDO5ADMVHB/events.json","paper":"https://pith.science/paper/7RU7FO3E"},"agent_actions":{"view_html":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB","download_json":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB.json","view_paper":"https://pith.science/paper/7RU7FO3E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.02775&json=true","fetch_graph":"https://pith.science/api/pith-number/7RU7FO3ERSFGWZ3LVDO5ADMVHB/graph.json","fetch_events":"https://pith.science/api/pith-number/7RU7FO3ERSFGWZ3LVDO5ADMVHB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB/action/storage_attestation","attest_author":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB/action/author_attestation","sign_citation":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB/action/citation_signature","submit_replication":"https://pith.science/pith/7RU7FO3ERSFGWZ3LVDO5ADMVHB/action/replication_record"}},"created_at":"2026-07-05T07:52:10.205926+00:00","updated_at":"2026-07-05T07:52:10.205926+00:00"}