{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XSIZ7SBGW5XSTDGWDLC6QALQUQ","short_pith_number":"pith:XSIZ7SBG","schema_version":"1.0","canonical_sha256":"bc919fc826b76f298cd61ac5e80170a426330a3322d53500f48df0d49a15f1e5","source":{"kind":"arxiv","id":"2504.14356","version":1},"attestation_state":"computed","paper":{"title":"Mathematical Programming Models for Exact and Interpretable Formulation of Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.AI","authors_text":"Edrin Hasaj, Jacob Gipp, Masoud Ataei, Sepideh Forouzi","submitted_at":"2025-04-19T16:50:17Z","abstract_excerpt":"This paper presents a unified mixed-integer programming framework for training sparse and interpretable neural networks. We develop exact formulations for both fully connected and convolutional architectures by modeling nonlinearities such as ReLU activations through binary variables and encoding structural sparsity via filter- and layer-level pruning constraints. The resulting models integrate parameter learning, architecture selection, and structural regularization within a single optimization problem, yielding globally optimal solutions with respect to a composite objective that balances pr"},"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":"2504.14356","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-19T16:50:17Z","cross_cats_sorted":["math.OC"],"title_canon_sha256":"298d992954ae6a0984d8d34dc6b0dacbbea9a889f0058c21100c06967b743823","abstract_canon_sha256":"f992df0e3c8b11047c08849fd802898f48e02df0f92b58b644a79bbdd0e7eb2c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:51:42.278941Z","signature_b64":"1qUIeKuI+Tt7e+synx8UHBJN9U9/Y1sUj9GBQRrQm1DKo7GnbLMo/62VTP4H0UtGRFzxqYMtpkJx0DHw0/sYAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bc919fc826b76f298cd61ac5e80170a426330a3322d53500f48df0d49a15f1e5","last_reissued_at":"2026-07-05T10:51:42.278440Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:51:42.278440Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mathematical Programming Models for Exact and Interpretable Formulation of Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC"],"primary_cat":"cs.AI","authors_text":"Edrin Hasaj, Jacob Gipp, Masoud Ataei, Sepideh Forouzi","submitted_at":"2025-04-19T16:50:17Z","abstract_excerpt":"This paper presents a unified mixed-integer programming framework for training sparse and interpretable neural networks. We develop exact formulations for both fully connected and convolutional architectures by modeling nonlinearities such as ReLU activations through binary variables and encoding structural sparsity via filter- and layer-level pruning constraints. The resulting models integrate parameter learning, architecture selection, and structural regularization within a single optimization problem, yielding globally optimal solutions with respect to a composite objective that balances pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.14356","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/2504.14356/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":"2504.14356","created_at":"2026-07-05T10:51:42.278502+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.14356v1","created_at":"2026-07-05T10:51:42.278502+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.14356","created_at":"2026-07-05T10:51:42.278502+00:00"},{"alias_kind":"pith_short_12","alias_value":"XSIZ7SBGW5XS","created_at":"2026-07-05T10:51:42.278502+00:00"},{"alias_kind":"pith_short_16","alias_value":"XSIZ7SBGW5XSTDGW","created_at":"2026-07-05T10:51:42.278502+00:00"},{"alias_kind":"pith_short_8","alias_value":"XSIZ7SBG","created_at":"2026-07-05T10:51:42.278502+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.16426","citing_title":"Functional Similarity Metric for Neural Networks: Overcoming Parametric Ambiguity via Activation Region Analysis","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ","json":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ.json","graph_json":"https://pith.science/api/pith-number/XSIZ7SBGW5XSTDGWDLC6QALQUQ/graph.json","events_json":"https://pith.science/api/pith-number/XSIZ7SBGW5XSTDGWDLC6QALQUQ/events.json","paper":"https://pith.science/paper/XSIZ7SBG"},"agent_actions":{"view_html":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ","download_json":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ.json","view_paper":"https://pith.science/paper/XSIZ7SBG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.14356&json=true","fetch_graph":"https://pith.science/api/pith-number/XSIZ7SBGW5XSTDGWDLC6QALQUQ/graph.json","fetch_events":"https://pith.science/api/pith-number/XSIZ7SBGW5XSTDGWDLC6QALQUQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ/action/storage_attestation","attest_author":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ/action/author_attestation","sign_citation":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ/action/citation_signature","submit_replication":"https://pith.science/pith/XSIZ7SBGW5XSTDGWDLC6QALQUQ/action/replication_record"}},"created_at":"2026-07-05T10:51:42.278502+00:00","updated_at":"2026-07-05T10:51:42.278502+00:00"}