{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:QCMI3THHLAXHYNADE56W37MK5C","short_pith_number":"pith:QCMI3THH","schema_version":"1.0","canonical_sha256":"80988dcce7582e7c3403277d6dfd8ae89493f5ca39645bd3c91aa9dcadcd3639","source":{"kind":"arxiv","id":"1704.00109","version":1},"attestation_state":"computed","paper":{"title":"Snapshot Ensembles: Train 1, get M for free","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Gao Huang, Geoff Pleiss, John E. Hopcroft, Kilian Q. Weinberger, Yixuan Li, Zhuang Liu","submitted_at":"2017-04-01T02:42:55Z","abstract_excerpt":"Ensembles of neural networks are known to be much more robust and accurate than individual networks. However, training multiple deep networks for model averaging is computationally expensive. In this paper, we propose a method to obtain the seemingly contradictory goal of ensembling multiple neural networks at no additional training cost. We achieve this goal by training a single neural network, converging to several local minima along its optimization path and saving the model parameters. To obtain repeated rapid convergence, we leverage recent work on cyclic learning rate schedules. The resu"},"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":"1704.00109","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-04-01T02:42:55Z","cross_cats_sorted":[],"title_canon_sha256":"7b9b7b651c70af06996438c3abd4174713c85e6af812b2fa47b4bad1477ef5ff","abstract_canon_sha256":"3089fee70881f344e67be712816ffb2c9a74173993205cf528924acdf04ebb89"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:47:25.427512Z","signature_b64":"mQaeaecKdtmwYd+pN8Z0/UjHiKZMw6s0p6iwADUkxtxR6jQB504CxF8yQQIngjmqcq5Mcpg01VTlrLCAeknPAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80988dcce7582e7c3403277d6dfd8ae89493f5ca39645bd3c91aa9dcadcd3639","last_reissued_at":"2026-05-18T00:47:25.426835Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:47:25.426835Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Snapshot Ensembles: Train 1, get M for free","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Gao Huang, Geoff Pleiss, John E. Hopcroft, Kilian Q. Weinberger, Yixuan Li, Zhuang Liu","submitted_at":"2017-04-01T02:42:55Z","abstract_excerpt":"Ensembles of neural networks are known to be much more robust and accurate than individual networks. However, training multiple deep networks for model averaging is computationally expensive. In this paper, we propose a method to obtain the seemingly contradictory goal of ensembling multiple neural networks at no additional training cost. We achieve this goal by training a single neural network, converging to several local minima along its optimization path and saving the model parameters. To obtain repeated rapid convergence, we leverage recent work on cyclic learning rate schedules. The resu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1704.00109","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":""},"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":"1704.00109","created_at":"2026-05-18T00:47:25.426932+00:00"},{"alias_kind":"arxiv_version","alias_value":"1704.00109v1","created_at":"2026-05-18T00:47:25.426932+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1704.00109","created_at":"2026-05-18T00:47:25.426932+00:00"},{"alias_kind":"pith_short_12","alias_value":"QCMI3THHLAXH","created_at":"2026-05-18T12:31:37.085036+00:00"},{"alias_kind":"pith_short_16","alias_value":"QCMI3THHLAXHYNAD","created_at":"2026-05-18T12:31:37.085036+00:00"},{"alias_kind":"pith_short_8","alias_value":"QCMI3THH","created_at":"2026-05-18T12:31:37.085036+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"1907.00350","citing_title":"Random Vector Functional Link Neural Network based Ensemble Deep Learning","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"1907.10418","citing_title":"Improving Malaria Parasite Detection from Red Blood Cell using Deep Convolutional Neural Networks","ref_index":58,"is_internal_anchor":true},{"citing_arxiv_id":"2405.07987","citing_title":"The Platonic Representation Hypothesis","ref_index":100,"is_internal_anchor":true},{"citing_arxiv_id":"2604.05500","citing_title":"CLIP-Guided Data Augmentation for Night-Time Image Dehazing","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C","json":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C.json","graph_json":"https://pith.science/api/pith-number/QCMI3THHLAXHYNADE56W37MK5C/graph.json","events_json":"https://pith.science/api/pith-number/QCMI3THHLAXHYNADE56W37MK5C/events.json","paper":"https://pith.science/paper/QCMI3THH"},"agent_actions":{"view_html":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C","download_json":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C.json","view_paper":"https://pith.science/paper/QCMI3THH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1704.00109&json=true","fetch_graph":"https://pith.science/api/pith-number/QCMI3THHLAXHYNADE56W37MK5C/graph.json","fetch_events":"https://pith.science/api/pith-number/QCMI3THHLAXHYNADE56W37MK5C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C/action/storage_attestation","attest_author":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C/action/author_attestation","sign_citation":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C/action/citation_signature","submit_replication":"https://pith.science/pith/QCMI3THHLAXHYNADE56W37MK5C/action/replication_record"}},"created_at":"2026-05-18T00:47:25.426932+00:00","updated_at":"2026-05-18T00:47:25.426932+00:00"}