{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:3OUT4QWSOLTQ7QI3RGNTQIFUD7","short_pith_number":"pith:3OUT4QWS","schema_version":"1.0","canonical_sha256":"dba93e42d272e70fc11b899b3820b41fcb1c85d11f63347152f122215ddd6705","source":{"kind":"arxiv","id":"2010.06610","version":2},"attestation_state":"computed","paper":{"title":"Training independent subnetworks for robust prediction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrew M. Dai, Balaji Lakshminarayanan, Dustin Tran, Jasper Snoek, Jeremiah Zhe Liu, Marton Havasi, Rodolphe Jenatton, Stanislav Fort","submitted_at":"2020-10-13T18:05:13Z","abstract_excerpt":"Recent approaches to efficiently ensemble neural networks have shown that strong robustness and uncertainty performance can be achieved with a negligible gain in parameters over the original network. However, these methods still require multiple forward passes for prediction, leading to a significant computational cost. In this work, we show a surprising result: the benefits of using multiple predictions can be achieved `for free' under a single model's forward pass. In particular, we show that, using a multi-input multi-output (MIMO) configuration, one can utilize a single model's capacity to"},"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":"2010.06610","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-10-13T18:05:13Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"9f4f2b0050bcf407e53ef3e8dff1ea5f6a4a9458ae608fd234ed77a23a1a6b9b","abstract_canon_sha256":"be2862211077abfe76f6fee05a5f06145a5964524de12f15293db0c0a560def8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:03:25.981915Z","signature_b64":"yOkI38Vl84+KYTBfTGPz+9+HsIFnYT59/huH/EJV5uu08eSbiCSnnYAAhWbx4h+XtBQMyNuwl6+VAaeUKc/GAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dba93e42d272e70fc11b899b3820b41fcb1c85d11f63347152f122215ddd6705","last_reissued_at":"2026-07-05T03:03:25.981493Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:03:25.981493Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training independent subnetworks for robust prediction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrew M. Dai, Balaji Lakshminarayanan, Dustin Tran, Jasper Snoek, Jeremiah Zhe Liu, Marton Havasi, Rodolphe Jenatton, Stanislav Fort","submitted_at":"2020-10-13T18:05:13Z","abstract_excerpt":"Recent approaches to efficiently ensemble neural networks have shown that strong robustness and uncertainty performance can be achieved with a negligible gain in parameters over the original network. However, these methods still require multiple forward passes for prediction, leading to a significant computational cost. In this work, we show a surprising result: the benefits of using multiple predictions can be achieved `for free' under a single model's forward pass. In particular, we show that, using a multi-input multi-output (MIMO) configuration, one can utilize a single model's capacity to"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.06610","kind":"arxiv","version":2},"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/2010.06610/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":"2010.06610","created_at":"2026-07-05T03:03:25.981554+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.06610v2","created_at":"2026-07-05T03:03:25.981554+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.06610","created_at":"2026-07-05T03:03:25.981554+00:00"},{"alias_kind":"pith_short_12","alias_value":"3OUT4QWSOLTQ","created_at":"2026-07-05T03:03:25.981554+00:00"},{"alias_kind":"pith_short_16","alias_value":"3OUT4QWSOLTQ7QI3","created_at":"2026-07-05T03:03:25.981554+00:00"},{"alias_kind":"pith_short_8","alias_value":"3OUT4QWS","created_at":"2026-07-05T03:03:25.981554+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09350","citing_title":"Taming Perception Jitter: Uncertainty-Aware LiDAR Object Detection for Reliable Motion Classification","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7","json":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7.json","graph_json":"https://pith.science/api/pith-number/3OUT4QWSOLTQ7QI3RGNTQIFUD7/graph.json","events_json":"https://pith.science/api/pith-number/3OUT4QWSOLTQ7QI3RGNTQIFUD7/events.json","paper":"https://pith.science/paper/3OUT4QWS"},"agent_actions":{"view_html":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7","download_json":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7.json","view_paper":"https://pith.science/paper/3OUT4QWS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.06610&json=true","fetch_graph":"https://pith.science/api/pith-number/3OUT4QWSOLTQ7QI3RGNTQIFUD7/graph.json","fetch_events":"https://pith.science/api/pith-number/3OUT4QWSOLTQ7QI3RGNTQIFUD7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7/action/storage_attestation","attest_author":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7/action/author_attestation","sign_citation":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7/action/citation_signature","submit_replication":"https://pith.science/pith/3OUT4QWSOLTQ7QI3RGNTQIFUD7/action/replication_record"}},"created_at":"2026-07-05T03:03:25.981554+00:00","updated_at":"2026-07-05T03:03:25.981554+00:00"}