{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:7MPDYKOJQ76YICJENCJ5LGGAOG","short_pith_number":"pith:7MPDYKOJ","schema_version":"1.0","canonical_sha256":"fb1e3c29c987fd8409246893d598c071a86f6d30192985e24711c1f73694a213","source":{"kind":"arxiv","id":"2509.00924","version":1},"attestation_state":"computed","paper":{"title":"Beyond Universal Approximation Theorems: Algorithmic Uniform Approximation by Neural Networks Trained with Noisy Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.NA","cs.NE","math.NA","math.PR"],"primary_cat":"stat.ML","authors_text":"Anastasis Kratsios, Daniel Roy, Tin Sum Cheng","submitted_at":"2025-08-31T16:20:27Z","abstract_excerpt":"At its core, machine learning seeks to train models that reliably generalize beyond noisy observations; however, the theoretical vacuum in which state-of-the-art universal approximation theorems (UATs) operate isolates them from this goal, as they assume noiseless data and allow network parameters to be chosen freely, independent of algorithmic realism. This paper bridges that gap by introducing an architecture-specific randomized training algorithm that constructs a uniform approximator from $N$ noisy training samples on the $d$-dimensional cube $[0,1]^d$. Our trained neural networks attain t"},"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":"2509.00924","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-08-31T16:20:27Z","cross_cats_sorted":["cs.LG","cs.NA","cs.NE","math.NA","math.PR"],"title_canon_sha256":"7a310d708c03061762dcd9392906f1afef80758260db0ce6d8f5ac0d37bdc281","abstract_canon_sha256":"1fce652a97325c44e95343874d1cf9d7df791cb739663de510b6d181b30aba24"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:02:24.726300Z","signature_b64":"t2tum4Vm+RDQLIHFz6I6E7H8+etlXlZcC8jze9ujj/kFgfdg9V7mJ2ty2Te3vFft9rPpqozShNAw6bU9TZQBAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fb1e3c29c987fd8409246893d598c071a86f6d30192985e24711c1f73694a213","last_reissued_at":"2026-07-05T12:02:24.725788Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:02:24.725788Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Universal Approximation Theorems: Algorithmic Uniform Approximation by Neural Networks Trained with Noisy Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.NA","cs.NE","math.NA","math.PR"],"primary_cat":"stat.ML","authors_text":"Anastasis Kratsios, Daniel Roy, Tin Sum Cheng","submitted_at":"2025-08-31T16:20:27Z","abstract_excerpt":"At its core, machine learning seeks to train models that reliably generalize beyond noisy observations; however, the theoretical vacuum in which state-of-the-art universal approximation theorems (UATs) operate isolates them from this goal, as they assume noiseless data and allow network parameters to be chosen freely, independent of algorithmic realism. This paper bridges that gap by introducing an architecture-specific randomized training algorithm that constructs a uniform approximator from $N$ noisy training samples on the $d$-dimensional cube $[0,1]^d$. Our trained neural networks attain t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.00924","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/2509.00924/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":"2509.00924","created_at":"2026-07-05T12:02:24.725848+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.00924v1","created_at":"2026-07-05T12:02:24.725848+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.00924","created_at":"2026-07-05T12:02:24.725848+00:00"},{"alias_kind":"pith_short_12","alias_value":"7MPDYKOJQ76Y","created_at":"2026-07-05T12:02:24.725848+00:00"},{"alias_kind":"pith_short_16","alias_value":"7MPDYKOJQ76YICJE","created_at":"2026-07-05T12:02:24.725848+00:00"},{"alias_kind":"pith_short_8","alias_value":"7MPDYKOJ","created_at":"2026-07-05T12:02:24.725848+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.04995","citing_title":"Adaptivity Under Realizability Constraints: Comparing In-Context and Agentic Learning","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG","json":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG.json","graph_json":"https://pith.science/api/pith-number/7MPDYKOJQ76YICJENCJ5LGGAOG/graph.json","events_json":"https://pith.science/api/pith-number/7MPDYKOJQ76YICJENCJ5LGGAOG/events.json","paper":"https://pith.science/paper/7MPDYKOJ"},"agent_actions":{"view_html":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG","download_json":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG.json","view_paper":"https://pith.science/paper/7MPDYKOJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.00924&json=true","fetch_graph":"https://pith.science/api/pith-number/7MPDYKOJQ76YICJENCJ5LGGAOG/graph.json","fetch_events":"https://pith.science/api/pith-number/7MPDYKOJQ76YICJENCJ5LGGAOG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG/action/storage_attestation","attest_author":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG/action/author_attestation","sign_citation":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG/action/citation_signature","submit_replication":"https://pith.science/pith/7MPDYKOJQ76YICJENCJ5LGGAOG/action/replication_record"}},"created_at":"2026-07-05T12:02:24.725848+00:00","updated_at":"2026-07-05T12:02:24.725848+00:00"}