{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:WJK4YZB3SGW6JOXCZOOTQBL7PF","short_pith_number":"pith:WJK4YZB3","schema_version":"1.0","canonical_sha256":"b255cc643b91ade4bae2cb9d38057f795f478b7004934a2d8cf33a89c1e63818","source":{"kind":"arxiv","id":"2209.15106","version":1},"attestation_state":"computed","paper":{"title":"Restricted Strong Convexity of Deep Learning Models with Smooth Activations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Arindam Banerjee, Libin Zhu, Mikhail Belkin, Pedro Cisneros-Velarde","submitted_at":"2022-09-29T21:24:26Z","abstract_excerpt":"We consider the problem of optimization of deep learning models with smooth activation functions. While there exist influential results on the problem from the ``near initialization'' perspective, we shed considerable new light on the problem. In particular, we make two key technical contributions for such models with $L$ layers, $m$ width, and $\\sigma_0^2$ initialization variance. First, for suitable $\\sigma_0^2$, we establish a $O(\\frac{\\text{poly}(L)}{\\sqrt{m}})$ upper bound on the spectral norm of the Hessian of such models, considerably sharpening prior results. Second, we introduce a new"},"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":"2209.15106","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-09-29T21:24:26Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"c919e8c5b5caeef5c8d9de498e5cebd4d726fbdff54f557a5c65e14de240dbc1","abstract_canon_sha256":"ca060670ac39e847d9ded0777badb81d073d50dc45601b442893e167ef215df2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:02:15.909349Z","signature_b64":"C8Yx9aOp43drbbVTSXE3GpNPcFsaLw8xaQAl0FylWFOLYfQq3jE6oDLTMsV4pTi3byurhIDkURwoKwbiz+T9Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b255cc643b91ade4bae2cb9d38057f795f478b7004934a2d8cf33a89c1e63818","last_reissued_at":"2026-07-05T05:02:15.908939Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:02:15.908939Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Restricted Strong Convexity of Deep Learning Models with Smooth Activations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.LG","authors_text":"Arindam Banerjee, Libin Zhu, Mikhail Belkin, Pedro Cisneros-Velarde","submitted_at":"2022-09-29T21:24:26Z","abstract_excerpt":"We consider the problem of optimization of deep learning models with smooth activation functions. While there exist influential results on the problem from the ``near initialization'' perspective, we shed considerable new light on the problem. In particular, we make two key technical contributions for such models with $L$ layers, $m$ width, and $\\sigma_0^2$ initialization variance. First, for suitable $\\sigma_0^2$, we establish a $O(\\frac{\\text{poly}(L)}{\\sqrt{m}})$ upper bound on the spectral norm of the Hessian of such models, considerably sharpening prior results. Second, we introduce a new"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.15106","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2209.15106/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":"2209.15106","created_at":"2026-07-05T05:02:15.908994+00:00"},{"alias_kind":"arxiv_version","alias_value":"2209.15106v1","created_at":"2026-07-05T05:02:15.908994+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.15106","created_at":"2026-07-05T05:02:15.908994+00:00"},{"alias_kind":"pith_short_12","alias_value":"WJK4YZB3SGW6","created_at":"2026-07-05T05:02:15.908994+00:00"},{"alias_kind":"pith_short_16","alias_value":"WJK4YZB3SGW6JOXC","created_at":"2026-07-05T05:02:15.908994+00:00"},{"alias_kind":"pith_short_8","alias_value":"WJK4YZB3","created_at":"2026-07-05T05:02:15.908994+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26749","citing_title":"Structure Before Collapse: Transient semantic geometry in next-token prediction","ref_index":203,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF","json":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF.json","graph_json":"https://pith.science/api/pith-number/WJK4YZB3SGW6JOXCZOOTQBL7PF/graph.json","events_json":"https://pith.science/api/pith-number/WJK4YZB3SGW6JOXCZOOTQBL7PF/events.json","paper":"https://pith.science/paper/WJK4YZB3"},"agent_actions":{"view_html":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF","download_json":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF.json","view_paper":"https://pith.science/paper/WJK4YZB3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2209.15106&json=true","fetch_graph":"https://pith.science/api/pith-number/WJK4YZB3SGW6JOXCZOOTQBL7PF/graph.json","fetch_events":"https://pith.science/api/pith-number/WJK4YZB3SGW6JOXCZOOTQBL7PF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF/action/storage_attestation","attest_author":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF/action/author_attestation","sign_citation":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF/action/citation_signature","submit_replication":"https://pith.science/pith/WJK4YZB3SGW6JOXCZOOTQBL7PF/action/replication_record"}},"created_at":"2026-07-05T05:02:15.908994+00:00","updated_at":"2026-07-05T05:02:15.908994+00:00"}