{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6NESL4CKOVMEYDPEQSNNMTKIZP","short_pith_number":"pith:6NESL4CK","schema_version":"1.0","canonical_sha256":"f34925f04a75584c0de4849ad64d48cbe5a4c9836af9a6ee42234be6ed85875f","source":{"kind":"arxiv","id":"2405.16630","version":1},"attestation_state":"computed","paper":{"title":"Bayesian Inference with Deep Weakly Nonlinear Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","math.PR","physics.data-an"],"primary_cat":"stat.ML","authors_text":"Alexander Zlokapa, Boris Hanin","submitted_at":"2024-05-26T17:08:04Z","abstract_excerpt":"We show at a physics level of rigor that Bayesian inference with a fully connected neural network and a shaped nonlinearity of the form $\\phi(t) = t + \\psi t^3/L$ is (perturbatively) solvable in the regime where the number of training datapoints $P$ , the input dimension $N_0$, the network layer widths $N$, and the network depth $L$ are simultaneously large. Our results hold with weak assumptions on the data; the main constraint is that $P < N_0$. We provide techniques to compute the model evidence and posterior to arbitrary order in $1/N$ and at arbitrary temperature. We report the following "},"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":"2405.16630","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-05-26T17:08:04Z","cross_cats_sorted":["cs.AI","cs.LG","math.PR","physics.data-an"],"title_canon_sha256":"201827fd4103dbb52ddd8de23f110e059211a8216d577d3bf4f518263eeab608","abstract_canon_sha256":"3bfe289cf366a57c3883b97ae3b52661f7f781d5c5d3e29977471ce24a019a88"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:23:33.437413Z","signature_b64":"qvauttGXOKRAa+T1Q0aWJA+MWzg1vKWrcqbM/IV6r+PifBUL9vKfni8QLfg4MPzlFvU6W/ffM/pflxWaDoeACg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f34925f04a75584c0de4849ad64d48cbe5a4c9836af9a6ee42234be6ed85875f","last_reissued_at":"2026-07-05T08:23:33.436960Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:23:33.436960Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bayesian Inference with Deep Weakly Nonlinear Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","math.PR","physics.data-an"],"primary_cat":"stat.ML","authors_text":"Alexander Zlokapa, Boris Hanin","submitted_at":"2024-05-26T17:08:04Z","abstract_excerpt":"We show at a physics level of rigor that Bayesian inference with a fully connected neural network and a shaped nonlinearity of the form $\\phi(t) = t + \\psi t^3/L$ is (perturbatively) solvable in the regime where the number of training datapoints $P$ , the input dimension $N_0$, the network layer widths $N$, and the network depth $L$ are simultaneously large. Our results hold with weak assumptions on the data; the main constraint is that $P < N_0$. We provide techniques to compute the model evidence and posterior to arbitrary order in $1/N$ and at arbitrary temperature. We report the following "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.16630","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/2405.16630/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":"2405.16630","created_at":"2026-07-05T08:23:33.437019+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.16630v1","created_at":"2026-07-05T08:23:33.437019+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.16630","created_at":"2026-07-05T08:23:33.437019+00:00"},{"alias_kind":"pith_short_12","alias_value":"6NESL4CKOVME","created_at":"2026-07-05T08:23:33.437019+00:00"},{"alias_kind":"pith_short_16","alias_value":"6NESL4CKOVMEYDPE","created_at":"2026-07-05T08:23:33.437019+00:00"},{"alias_kind":"pith_short_8","alias_value":"6NESL4CK","created_at":"2026-07-05T08:23:33.437019+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20299","citing_title":"Statistical Properties of Training & Generalization","ref_index":185,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20299","citing_title":"Statistical Properties of Training & Generalization","ref_index":185,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30831","citing_title":"Geometric Dyson Brownian Motions and the Free Log-Normal Limit for a Non-Square Gaussian Matrix Product","ref_index":59,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP","json":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP.json","graph_json":"https://pith.science/api/pith-number/6NESL4CKOVMEYDPEQSNNMTKIZP/graph.json","events_json":"https://pith.science/api/pith-number/6NESL4CKOVMEYDPEQSNNMTKIZP/events.json","paper":"https://pith.science/paper/6NESL4CK"},"agent_actions":{"view_html":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP","download_json":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP.json","view_paper":"https://pith.science/paper/6NESL4CK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.16630&json=true","fetch_graph":"https://pith.science/api/pith-number/6NESL4CKOVMEYDPEQSNNMTKIZP/graph.json","fetch_events":"https://pith.science/api/pith-number/6NESL4CKOVMEYDPEQSNNMTKIZP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP/action/storage_attestation","attest_author":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP/action/author_attestation","sign_citation":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP/action/citation_signature","submit_replication":"https://pith.science/pith/6NESL4CKOVMEYDPEQSNNMTKIZP/action/replication_record"}},"created_at":"2026-07-05T08:23:33.437019+00:00","updated_at":"2026-07-05T08:23:33.437019+00:00"}