{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:56STNILHMD5HPPSV76Z5NUJSWM","short_pith_number":"pith:56STNILH","schema_version":"1.0","canonical_sha256":"efa536a16760fa77be55ffb3d6d132b3019645877c39100b8c041e39d35d7705","source":{"kind":"arxiv","id":"2205.14189","version":2},"attestation_state":"computed","paper":{"title":"Optimizing Objective Functions from Trained ReLU Neural Networks via Sampling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"math.OC","authors_text":"Asterios Tsiourvas, Georgia Perakis","submitted_at":"2022-05-27T18:35:48Z","abstract_excerpt":"This paper introduces scalable, sampling-based algorithms that optimize trained neural networks with ReLU activations. We first propose an iterative algorithm that takes advantage of the piecewise linear structure of ReLU neural networks and reduces the initial mixed-integer optimization problem (MIP) into multiple easy-to-solve linear optimization problems (LPs) through sampling. Subsequently, we extend this approach by searching around the neighborhood of the LP solution computed at each iteration. This scheme allows us to devise a second, enhanced algorithm that reduces the initial MIP prob"},"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":"2205.14189","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"math.OC","submitted_at":"2022-05-27T18:35:48Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"3775b04e4d461e170f2f4ff4ffbe8f12ea37bfc0de60a1482fe3ab59bbd140b4","abstract_canon_sha256":"36d3c399201d42c75e0f8783344128c1611f46b047aa456e1c17ae91a0c1a4a7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:29:13.950334Z","signature_b64":"meUprh1csgWZRM2lJQY+zdmkKGzqnaSdORzR7wVq216h+xddbBW2regjwtYSVtu0JHaPlAg7/R6w87MGxG9kAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"efa536a16760fa77be55ffb3d6d132b3019645877c39100b8c041e39d35d7705","last_reissued_at":"2026-07-05T04:29:13.949887Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:29:13.949887Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Optimizing Objective Functions from Trained ReLU Neural Networks via Sampling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"math.OC","authors_text":"Asterios Tsiourvas, Georgia Perakis","submitted_at":"2022-05-27T18:35:48Z","abstract_excerpt":"This paper introduces scalable, sampling-based algorithms that optimize trained neural networks with ReLU activations. We first propose an iterative algorithm that takes advantage of the piecewise linear structure of ReLU neural networks and reduces the initial mixed-integer optimization problem (MIP) into multiple easy-to-solve linear optimization problems (LPs) through sampling. Subsequently, we extend this approach by searching around the neighborhood of the LP solution computed at each iteration. This scheme allows us to devise a second, enhanced algorithm that reduces the initial MIP prob"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.14189","kind":"arxiv","version":2},"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/2205.14189/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":"2205.14189","created_at":"2026-07-05T04:29:13.949950+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.14189v2","created_at":"2026-07-05T04:29:13.949950+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.14189","created_at":"2026-07-05T04:29:13.949950+00:00"},{"alias_kind":"pith_short_12","alias_value":"56STNILHMD5H","created_at":"2026-07-05T04:29:13.949950+00:00"},{"alias_kind":"pith_short_16","alias_value":"56STNILHMD5HPPSV","created_at":"2026-07-05T04:29:13.949950+00:00"},{"alias_kind":"pith_short_8","alias_value":"56STNILH","created_at":"2026-07-05T04:29:13.949950+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.22746","citing_title":"Relaxation-Informed Training of Neural Network Surrogate Models","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM","json":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM.json","graph_json":"https://pith.science/api/pith-number/56STNILHMD5HPPSV76Z5NUJSWM/graph.json","events_json":"https://pith.science/api/pith-number/56STNILHMD5HPPSV76Z5NUJSWM/events.json","paper":"https://pith.science/paper/56STNILH"},"agent_actions":{"view_html":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM","download_json":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM.json","view_paper":"https://pith.science/paper/56STNILH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.14189&json=true","fetch_graph":"https://pith.science/api/pith-number/56STNILHMD5HPPSV76Z5NUJSWM/graph.json","fetch_events":"https://pith.science/api/pith-number/56STNILHMD5HPPSV76Z5NUJSWM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM/action/storage_attestation","attest_author":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM/action/author_attestation","sign_citation":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM/action/citation_signature","submit_replication":"https://pith.science/pith/56STNILHMD5HPPSV76Z5NUJSWM/action/replication_record"}},"created_at":"2026-07-05T04:29:13.949950+00:00","updated_at":"2026-07-05T04:29:13.949950+00:00"}