{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:UNWG4IEPL7W672U3Y7W4RIZRMD","short_pith_number":"pith:UNWG4IEP","schema_version":"1.0","canonical_sha256":"a36c6e208f5fedefea9bc7edc8a33160eeaa5d0b0150116660d17c753884143f","source":{"kind":"arxiv","id":"1805.04770","version":2},"attestation_state":"computed","paper":{"title":"Born Again Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"stat.ML","authors_text":"Anima Anandkumar, Laurent Itti, Michael Tschannen, Tommaso Furlanello, Zachary C. Lipton","submitted_at":"2018-05-12T19:48:50Z","abstract_excerpt":"Knowledge Distillation (KD) consists of transferring “knowledge” from one machine learning model (the teacher) to another (the student). Commonly, the teacher is a high-capacity model with formidable performance, while the student is more compact. By transferring knowledge, one hopes to benefit from the student’s compactness, without sacrificing too much performance. We study KD from a new perspective: rather than compressing models, we train students parameterized identically to their teachers. Surprisingly, these Born-Again Networks (BANs), outperform their teachers significantly, both"},"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":"1805.04770","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-05-12T19:48:50Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"f92430057cc79dc5884e83bdeaf9f2d0fda305160f3c2dc3b8355b6a1d3ada0e","abstract_canon_sha256":"0d1a060652c2e3394356f9645f80bb189354fbed7c4e14e943de7592b2659578"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:51:21.807245Z","signature_b64":"dEEpdUIZahQ6DmF+UvOgvdVi5j2ApgD7Irxi3YQ6nWmy2n/ksd1KpId1q2R9LK6uu0eiXGGzo/9mRl2kfVQVDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a36c6e208f5fedefea9bc7edc8a33160eeaa5d0b0150116660d17c753884143f","last_reissued_at":"2026-07-05T07:51:21.806773Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:51:21.806773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Born Again Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"stat.ML","authors_text":"Anima Anandkumar, Laurent Itti, Michael Tschannen, Tommaso Furlanello, Zachary C. Lipton","submitted_at":"2018-05-12T19:48:50Z","abstract_excerpt":"Knowledge Distillation (KD) consists of transferring “knowledge” from one machine learning model (the teacher) to another (the student). Commonly, the teacher is a high-capacity model with formidable performance, while the student is more compact. By transferring knowledge, one hopes to benefit from the student’s compactness, without sacrificing too much performance. We study KD from a new perspective: rather than compressing models, we train students parameterized identically to their teachers. Surprisingly, these Born-Again Networks (BANs), outperform their teachers significantly, both"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1805.04770","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/1805.04770/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":"1805.04770","created_at":"2026-07-05T07:51:21.806832+00:00"},{"alias_kind":"arxiv_version","alias_value":"1805.04770v2","created_at":"2026-07-05T07:51:21.806832+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1805.04770","created_at":"2026-07-05T07:51:21.806832+00:00"},{"alias_kind":"pith_short_12","alias_value":"UNWG4IEPL7W6","created_at":"2026-07-05T07:51:21.806832+00:00"},{"alias_kind":"pith_short_16","alias_value":"UNWG4IEPL7W672U3","created_at":"2026-07-05T07:51:21.806832+00:00"},{"alias_kind":"pith_short_8","alias_value":"UNWG4IEP","created_at":"2026-07-05T07:51:21.806832+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00173","citing_title":"TallyTrain: Communication-Efficient Federated Distillation","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03938","citing_title":"q0: Primitives for Hyper-Epoch Pretraining","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"1907.03402","citing_title":"Distill-2MD-MTL: Data Distillation based on Multi-Dataset Multi-Domain Multi-Task Frame Work to Solve Face Related Tasksks, Multi Task Learning, Semi-Supervised Learning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2101.02388","citing_title":"Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD","json":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD.json","graph_json":"https://pith.science/api/pith-number/UNWG4IEPL7W672U3Y7W4RIZRMD/graph.json","events_json":"https://pith.science/api/pith-number/UNWG4IEPL7W672U3Y7W4RIZRMD/events.json","paper":"https://pith.science/paper/UNWG4IEP"},"agent_actions":{"view_html":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD","download_json":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD.json","view_paper":"https://pith.science/paper/UNWG4IEP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1805.04770&json=true","fetch_graph":"https://pith.science/api/pith-number/UNWG4IEPL7W672U3Y7W4RIZRMD/graph.json","fetch_events":"https://pith.science/api/pith-number/UNWG4IEPL7W672U3Y7W4RIZRMD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD/action/storage_attestation","attest_author":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD/action/author_attestation","sign_citation":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD/action/citation_signature","submit_replication":"https://pith.science/pith/UNWG4IEPL7W672U3Y7W4RIZRMD/action/replication_record"}},"created_at":"2026-07-05T07:51:21.806832+00:00","updated_at":"2026-07-05T07:51:21.806832+00:00"}