{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:62WSKAANENLLVFXVL67TOSPXLM","short_pith_number":"pith:62WSKAAN","schema_version":"1.0","canonical_sha256":"f6ad25000d2356ba96f55fbf3749f75b2c57469046ad9e1fc8ef3fc35002c99d","source":{"kind":"arxiv","id":"2505.09167","version":1},"attestation_state":"computed","paper":{"title":"Online Learning of Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Amit Daniely, Elchanan Mossel, Idan Mehalel","submitted_at":"2025-05-14T06:03:07Z","abstract_excerpt":"We study online learning of feedforward neural networks with the sign activation function that implement functions from the unit ball in $\\mathbb{R}^d$ to a finite label set $\\{1, \\ldots, Y\\}$.\n  First, we characterize a margin condition that is sufficient and in some cases necessary for online learnability of a neural network: Every neuron in the first hidden layer classifies all instances with some margin $\\gamma$ bounded away from zero. Quantitatively, we prove that for any net, the optimal mistake bound is at most approximately $\\mathtt{TS}(d,\\gamma)$, which is the $(d,\\gamma)$-totally-sep"},"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":"2505.09167","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-05-14T06:03:07Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3f9ac6fe3efe6793ab49708cbbf81ad5ce25189243c5e652f30833d031171512","abstract_canon_sha256":"fd9ad5c9919c557cb148cb7ee384d365b4940ba49f7eb14b01fd272b80aedb91"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:02:51.422190Z","signature_b64":"NA88a+q9e0b89ulVO1yHld7sFrikrTuNfAr03Y2YdwkSN2a7LZCPwSYE/VFsAv9ez0YUpb+aEiqDvtrXZOzYDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f6ad25000d2356ba96f55fbf3749f75b2c57469046ad9e1fc8ef3fc35002c99d","last_reissued_at":"2026-07-05T11:02:51.421662Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:02:51.421662Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Online Learning of Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Amit Daniely, Elchanan Mossel, Idan Mehalel","submitted_at":"2025-05-14T06:03:07Z","abstract_excerpt":"We study online learning of feedforward neural networks with the sign activation function that implement functions from the unit ball in $\\mathbb{R}^d$ to a finite label set $\\{1, \\ldots, Y\\}$.\n  First, we characterize a margin condition that is sufficient and in some cases necessary for online learnability of a neural network: Every neuron in the first hidden layer classifies all instances with some margin $\\gamma$ bounded away from zero. Quantitatively, we prove that for any net, the optimal mistake bound is at most approximately $\\mathtt{TS}(d,\\gamma)$, which is the $(d,\\gamma)$-totally-sep"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.09167","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/2505.09167/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":"2505.09167","created_at":"2026-07-05T11:02:51.421717+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.09167v1","created_at":"2026-07-05T11:02:51.421717+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.09167","created_at":"2026-07-05T11:02:51.421717+00:00"},{"alias_kind":"pith_short_12","alias_value":"62WSKAANENLL","created_at":"2026-07-05T11:02:51.421717+00:00"},{"alias_kind":"pith_short_16","alias_value":"62WSKAANENLLVFXV","created_at":"2026-07-05T11:02:51.421717+00:00"},{"alias_kind":"pith_short_8","alias_value":"62WSKAAN","created_at":"2026-07-05T11:02:51.421717+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.12013","citing_title":"Sample Complexity of Autoregressive Reasoning: Chain-of-Thought vs. End-to-End","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM","json":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM.json","graph_json":"https://pith.science/api/pith-number/62WSKAANENLLVFXVL67TOSPXLM/graph.json","events_json":"https://pith.science/api/pith-number/62WSKAANENLLVFXVL67TOSPXLM/events.json","paper":"https://pith.science/paper/62WSKAAN"},"agent_actions":{"view_html":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM","download_json":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM.json","view_paper":"https://pith.science/paper/62WSKAAN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.09167&json=true","fetch_graph":"https://pith.science/api/pith-number/62WSKAANENLLVFXVL67TOSPXLM/graph.json","fetch_events":"https://pith.science/api/pith-number/62WSKAANENLLVFXVL67TOSPXLM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM/action/storage_attestation","attest_author":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM/action/author_attestation","sign_citation":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM/action/citation_signature","submit_replication":"https://pith.science/pith/62WSKAANENLLVFXVL67TOSPXLM/action/replication_record"}},"created_at":"2026-07-05T11:02:51.421717+00:00","updated_at":"2026-07-05T11:02:51.421717+00:00"}