{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:S23WYCUHZY3UDA4GEOROXDDTCZ","short_pith_number":"pith:S23WYCUH","schema_version":"1.0","canonical_sha256":"96b76c0a87ce3741838623a2eb8c73167fbc2b0e6ebcf97ecc715e8d0298334b","source":{"kind":"arxiv","id":"2505.11589","version":1},"attestation_state":"computed","paper":{"title":"A Training Framework for Optimal and Stable Training of Polynomial Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Forsad Al Hossain, Tauhidur Rahman","submitted_at":"2025-05-16T18:00:02Z","abstract_excerpt":"By replacing standard non-linearities with polynomial activations, Polynomial Neural Networks (PNNs) are pivotal for applications such as privacy-preserving inference via Homomorphic Encryption (HE). However, training PNNs effectively presents a significant challenge: low-degree polynomials can limit model expressivity, while higher-degree polynomials, crucial for capturing complex functions, often suffer from numerical instability and gradient explosion. We introduce a robust and versatile training framework featuring two synergistic innovations: 1) a novel Boundary Loss that exponentially pe"},"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":"2505.11589","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-16T18:00:02Z","cross_cats_sorted":[],"title_canon_sha256":"009260f4b09c42a74c380b092395ec4632fa54d8643d3d8c93c2d1ee8b5d99d3","abstract_canon_sha256":"1132509a8d3c92cd9fef8cf0d88d5b24009a4206ec3db1c938e73aaa596aa9af"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:04:24.121775Z","signature_b64":"IpKilba6gbxf23x4Y1/bW66vHiztes4AEx9NYqOrKmEL4CXubqmDojX8MBTVrovAmiPYMj89PUPjwQ04o7Z0Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"96b76c0a87ce3741838623a2eb8c73167fbc2b0e6ebcf97ecc715e8d0298334b","last_reissued_at":"2026-07-05T11:04:24.121231Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:04:24.121231Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Training Framework for Optimal and Stable Training of Polynomial Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Forsad Al Hossain, Tauhidur Rahman","submitted_at":"2025-05-16T18:00:02Z","abstract_excerpt":"By replacing standard non-linearities with polynomial activations, Polynomial Neural Networks (PNNs) are pivotal for applications such as privacy-preserving inference via Homomorphic Encryption (HE). However, training PNNs effectively presents a significant challenge: low-degree polynomials can limit model expressivity, while higher-degree polynomials, crucial for capturing complex functions, often suffer from numerical instability and gradient explosion. We introduce a robust and versatile training framework featuring two synergistic innovations: 1) a novel Boundary Loss that exponentially pe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11589","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/2505.11589/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":"2505.11589","created_at":"2026-07-05T11:04:24.121293+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.11589v1","created_at":"2026-07-05T11:04:24.121293+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11589","created_at":"2026-07-05T11:04:24.121293+00:00"},{"alias_kind":"pith_short_12","alias_value":"S23WYCUHZY3U","created_at":"2026-07-05T11:04:24.121293+00:00"},{"alias_kind":"pith_short_16","alias_value":"S23WYCUHZY3UDA4G","created_at":"2026-07-05T11:04:24.121293+00:00"},{"alias_kind":"pith_short_8","alias_value":"S23WYCUH","created_at":"2026-07-05T11:04:24.121293+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.09609","citing_title":"Minimal Filling Architectures of Polynomial Neural Networks: Counterexamples, Frontier Search, and Defects","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09609","citing_title":"Minimal Filling Architectures of Polynomial Neural Networks: Counterexamples, Frontier Search, and Defects","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ","json":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ.json","graph_json":"https://pith.science/api/pith-number/S23WYCUHZY3UDA4GEOROXDDTCZ/graph.json","events_json":"https://pith.science/api/pith-number/S23WYCUHZY3UDA4GEOROXDDTCZ/events.json","paper":"https://pith.science/paper/S23WYCUH"},"agent_actions":{"view_html":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ","download_json":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ.json","view_paper":"https://pith.science/paper/S23WYCUH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.11589&json=true","fetch_graph":"https://pith.science/api/pith-number/S23WYCUHZY3UDA4GEOROXDDTCZ/graph.json","fetch_events":"https://pith.science/api/pith-number/S23WYCUHZY3UDA4GEOROXDDTCZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ/action/storage_attestation","attest_author":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ/action/author_attestation","sign_citation":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ/action/citation_signature","submit_replication":"https://pith.science/pith/S23WYCUHZY3UDA4GEOROXDDTCZ/action/replication_record"}},"created_at":"2026-07-05T11:04:24.121293+00:00","updated_at":"2026-07-05T11:04:24.121293+00:00"}