{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NNHVNADTHTKQECTWZSDDG4MPHF","short_pith_number":"pith:NNHVNADT","schema_version":"1.0","canonical_sha256":"6b4f5680733cd5020a76cc8633718f396c54b8b0f1d242c1a92526458c2656f2","source":{"kind":"arxiv","id":"2509.11088","version":1},"attestation_state":"computed","paper":{"title":"Algebraic geometry of rational neural networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"math.AG","authors_text":"Alexandros Grosdos, Elina Robeva, Maksym Zubkov","submitted_at":"2025-09-14T04:34:21Z","abstract_excerpt":"We study the expressivity of rational neural networks (RationalNets) through the lens of algebraic geometry. We consider rational functions that arise from a given RationalNet to be tuples of fractions of homogeneous polynomials of fixed degrees. For a given architecture, the neuromanifold is the set of all such expressible tuples. For RationalNets with one hidden layer and fixed activation function $1/x$, we characterize the dimension of the neuromanifold and provide defining equations for some architectures. We also propose algorithms that determine whether a given rational function belongs "},"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":"2509.11088","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"math.AG","submitted_at":"2025-09-14T04:34:21Z","cross_cats_sorted":[],"title_canon_sha256":"ff5b7858238c5f930cce176c6ec4178738304c4b2811abd5ea5e3d9f4b13f336","abstract_canon_sha256":"bc2b68e9fbac9901ffb1f5f968fb1d120fddb7cc72c120dbed02507948df0130"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:11:33.322120Z","signature_b64":"SrZuHrV9zsYHgiPgLj6BO7qEDtaos+5Md44NEZf7CoULfwM8s6PY05OsnSle/u/J2LRJRVeqP4A8a6ZQDOCqAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b4f5680733cd5020a76cc8633718f396c54b8b0f1d242c1a92526458c2656f2","last_reissued_at":"2026-07-05T12:11:33.321522Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:11:33.321522Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Algebraic geometry of rational neural networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"math.AG","authors_text":"Alexandros Grosdos, Elina Robeva, Maksym Zubkov","submitted_at":"2025-09-14T04:34:21Z","abstract_excerpt":"We study the expressivity of rational neural networks (RationalNets) through the lens of algebraic geometry. We consider rational functions that arise from a given RationalNet to be tuples of fractions of homogeneous polynomials of fixed degrees. For a given architecture, the neuromanifold is the set of all such expressible tuples. For RationalNets with one hidden layer and fixed activation function $1/x$, we characterize the dimension of the neuromanifold and provide defining equations for some architectures. We also propose algorithms that determine whether a given rational function belongs "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.11088","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/2509.11088/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":"2509.11088","created_at":"2026-07-05T12:11:33.321596+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.11088v1","created_at":"2026-07-05T12:11:33.321596+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.11088","created_at":"2026-07-05T12:11:33.321596+00:00"},{"alias_kind":"pith_short_12","alias_value":"NNHVNADTHTKQ","created_at":"2026-07-05T12:11:33.321596+00:00"},{"alias_kind":"pith_short_16","alias_value":"NNHVNADTHTKQECTW","created_at":"2026-07-05T12:11:33.321596+00:00"},{"alias_kind":"pith_short_8","alias_value":"NNHVNADT","created_at":"2026-07-05T12:11:33.321596+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19268","citing_title":"Patnaik-Pearson intrinsic dimension for internal representations of neural networks","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19268","citing_title":"Patnaik-Pearson intrinsic dimension for internal representations of neural networks","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15970","citing_title":"Copositive Matrices with Ordered Off-Diagonal Entries","ref_index":137,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF","json":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF.json","graph_json":"https://pith.science/api/pith-number/NNHVNADTHTKQECTWZSDDG4MPHF/graph.json","events_json":"https://pith.science/api/pith-number/NNHVNADTHTKQECTWZSDDG4MPHF/events.json","paper":"https://pith.science/paper/NNHVNADT"},"agent_actions":{"view_html":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF","download_json":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF.json","view_paper":"https://pith.science/paper/NNHVNADT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.11088&json=true","fetch_graph":"https://pith.science/api/pith-number/NNHVNADTHTKQECTWZSDDG4MPHF/graph.json","fetch_events":"https://pith.science/api/pith-number/NNHVNADTHTKQECTWZSDDG4MPHF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF/action/storage_attestation","attest_author":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF/action/author_attestation","sign_citation":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF/action/citation_signature","submit_replication":"https://pith.science/pith/NNHVNADTHTKQECTWZSDDG4MPHF/action/replication_record"}},"created_at":"2026-07-05T12:11:33.321596+00:00","updated_at":"2026-07-05T12:11:33.321596+00:00"}