{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:YK5OFHTQETNZQYQWOVZFBY3Q65","short_pith_number":"pith:YK5OFHTQ","schema_version":"1.0","canonical_sha256":"c2bae29e7024db986216757250e370f752cd57a8d2157b4145c017dea56a4045","source":{"kind":"arxiv","id":"2205.08199","version":4},"attestation_state":"computed","paper":{"title":"Sharp asymptotics on the compression of two-layer neural networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.IT","stat.ML"],"primary_cat":"cs.IT","authors_text":"Marco Mondelli, Mohammad Hossein Amani, Rattana Pukdee, Simone Bombari, Stefano Rini","submitted_at":"2022-05-17T09:45:23Z","abstract_excerpt":"In this paper, we study the compression of a target two-layer neural network with N nodes into a compressed network with M<N nodes. More precisely, we consider the setting in which the weights of the target network are i.i.d. sub-Gaussian, and we minimize the population L_2 loss between the outputs of the target and of the compressed network, under the assumption of Gaussian inputs. By using tools from high-dimensional probability, we show that this non-convex problem can be simplified when the target network is sufficiently over-parameterized, and provide the error rate of this approximation "},"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":"2205.08199","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IT","submitted_at":"2022-05-17T09:45:23Z","cross_cats_sorted":["cs.LG","math.IT","stat.ML"],"title_canon_sha256":"7bc2f9acc96382222c6ac602927bf8fad84619e98d40db8e8f280aa03e804bf5","abstract_canon_sha256":"9f278b8dd281c152e5b31608c958cebf863af58f140eb0b5ed8c8d2a1a156db1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:48:52.445997Z","signature_b64":"dQfkQkVQp1ogCBnpCTUoH8ENVD7ubNIyr9HIPB7nv01TQ/doq3tKNWCH35G46l7lZ/4stgZ3Jnq9OXd3hNQ5Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c2bae29e7024db986216757250e370f752cd57a8d2157b4145c017dea56a4045","last_reissued_at":"2026-07-05T04:48:52.445603Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:48:52.445603Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sharp asymptotics on the compression of two-layer neural networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","math.IT","stat.ML"],"primary_cat":"cs.IT","authors_text":"Marco Mondelli, Mohammad Hossein Amani, Rattana Pukdee, Simone Bombari, Stefano Rini","submitted_at":"2022-05-17T09:45:23Z","abstract_excerpt":"In this paper, we study the compression of a target two-layer neural network with N nodes into a compressed network with M<N nodes. More precisely, we consider the setting in which the weights of the target network are i.i.d. sub-Gaussian, and we minimize the population L_2 loss between the outputs of the target and of the compressed network, under the assumption of Gaussian inputs. By using tools from high-dimensional probability, we show that this non-convex problem can be simplified when the target network is sufficiently over-parameterized, and provide the error rate of this approximation "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.08199","kind":"arxiv","version":4},"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/2205.08199/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":"2205.08199","created_at":"2026-07-05T04:48:52.445663+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.08199v4","created_at":"2026-07-05T04:48:52.445663+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.08199","created_at":"2026-07-05T04:48:52.445663+00:00"},{"alias_kind":"pith_short_12","alias_value":"YK5OFHTQETNZ","created_at":"2026-07-05T04:48:52.445663+00:00"},{"alias_kind":"pith_short_16","alias_value":"YK5OFHTQETNZQYQW","created_at":"2026-07-05T04:48:52.445663+00:00"},{"alias_kind":"pith_short_8","alias_value":"YK5OFHTQ","created_at":"2026-07-05T04:48:52.445663+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65","json":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65.json","graph_json":"https://pith.science/api/pith-number/YK5OFHTQETNZQYQWOVZFBY3Q65/graph.json","events_json":"https://pith.science/api/pith-number/YK5OFHTQETNZQYQWOVZFBY3Q65/events.json","paper":"https://pith.science/paper/YK5OFHTQ"},"agent_actions":{"view_html":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65","download_json":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65.json","view_paper":"https://pith.science/paper/YK5OFHTQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.08199&json=true","fetch_graph":"https://pith.science/api/pith-number/YK5OFHTQETNZQYQWOVZFBY3Q65/graph.json","fetch_events":"https://pith.science/api/pith-number/YK5OFHTQETNZQYQWOVZFBY3Q65/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65/action/storage_attestation","attest_author":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65/action/author_attestation","sign_citation":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65/action/citation_signature","submit_replication":"https://pith.science/pith/YK5OFHTQETNZQYQWOVZFBY3Q65/action/replication_record"}},"created_at":"2026-07-05T04:48:52.445663+00:00","updated_at":"2026-07-05T04:48:52.445663+00:00"}