{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:SUEVI5JLHLNPELCPTAMSIQ7PWM","short_pith_number":"pith:SUEVI5JL","schema_version":"1.0","canonical_sha256":"950954752b3adaf22c4f98192443efb318492885e400ae96562d61794f4d8697","source":{"kind":"arxiv","id":"2011.13824","version":2},"attestation_state":"computed","paper":{"title":"Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Cho-Jui Hsieh, Huan Zhang, Kaidi Xu, Shiqi Wang, Suman Jana, Xue Lin, Yihan Wang","submitted_at":"2020-11-27T16:42:12Z","abstract_excerpt":"Formal verification of neural networks (NNs) is a challenging and important problem. Existing efficient complete solvers typically require the branch-and-bound (BaB) process, which splits the problem domain into sub-domains and solves each sub-domain using faster but weaker incomplete verifiers, such as Linear Programming (LP) on linearly relaxed sub-domains. In this paper, we propose to use the backward mode linear relaxation based perturbation analysis (LiRPA) to replace LP during the BaB process, which can be efficiently implemented on the typical machine learning accelerators such as GPUs "},"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":"2011.13824","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2020-11-27T16:42:12Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"82a0ce815390ad371a2ab0998740dd567b45c7c7307b00e610399a4573d973c5","abstract_canon_sha256":"86db0a5cffff90b4918c28f1e10be83db59f4118eb6801a482adbf1c4fa410e3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:23:19.780425Z","signature_b64":"TmlzjeEMT+5yfrgNJ/eBDY+Rei+S+9kokLlFOxlY6WjGBZcy0xRUkWYEggf14L/+zEGTu9Y1nFdwImo0I6axCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"950954752b3adaf22c4f98192443efb318492885e400ae96562d61794f4d8697","last_reissued_at":"2026-07-05T02:23:19.779894Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:23:19.779894Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fast and Complete: Enabling Complete Neural Network Verification with Rapid and Massively Parallel Incomplete Verifiers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Cho-Jui Hsieh, Huan Zhang, Kaidi Xu, Shiqi Wang, Suman Jana, Xue Lin, Yihan Wang","submitted_at":"2020-11-27T16:42:12Z","abstract_excerpt":"Formal verification of neural networks (NNs) is a challenging and important problem. Existing efficient complete solvers typically require the branch-and-bound (BaB) process, which splits the problem domain into sub-domains and solves each sub-domain using faster but weaker incomplete verifiers, such as Linear Programming (LP) on linearly relaxed sub-domains. In this paper, we propose to use the backward mode linear relaxation based perturbation analysis (LiRPA) to replace LP during the BaB process, which can be efficiently implemented on the typical machine learning accelerators such as GPUs "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2011.13824","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/2011.13824/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":"2011.13824","created_at":"2026-07-05T02:23:19.779957+00:00"},{"alias_kind":"arxiv_version","alias_value":"2011.13824v2","created_at":"2026-07-05T02:23:19.779957+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2011.13824","created_at":"2026-07-05T02:23:19.779957+00:00"},{"alias_kind":"pith_short_12","alias_value":"SUEVI5JLHLNP","created_at":"2026-07-05T02:23:19.779957+00:00"},{"alias_kind":"pith_short_16","alias_value":"SUEVI5JLHLNPELCP","created_at":"2026-07-05T02:23:19.779957+00:00"},{"alias_kind":"pith_short_8","alias_value":"SUEVI5JL","created_at":"2026-07-05T02:23:19.779957+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.30155","citing_title":"Neural Network Verification using Partial Multi-Neuron Relaxation","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25239","citing_title":"Tensor-Based Batch Fuzzing with Adaptive Perturbation Scaling for Deep Neural Networks","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17153","citing_title":"Stress-Testing Neural Network Verifiers with Provably Robust Instances","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13845","citing_title":"Quantitative Linear Logic for Neuro-Symbolic Learning and Verification","ref_index":130,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13845","citing_title":"Quantitative Linear Logic for Neuro-Symbolic Learning and Verification","ref_index":130,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10621","citing_title":"Hierarchical End-to-End Taylor Bounds for Complete Neural Network Verification","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23474","citing_title":"GeoCert: Certified Geometric AI for Reliable Forecasting","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM","json":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM.json","graph_json":"https://pith.science/api/pith-number/SUEVI5JLHLNPELCPTAMSIQ7PWM/graph.json","events_json":"https://pith.science/api/pith-number/SUEVI5JLHLNPELCPTAMSIQ7PWM/events.json","paper":"https://pith.science/paper/SUEVI5JL"},"agent_actions":{"view_html":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM","download_json":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM.json","view_paper":"https://pith.science/paper/SUEVI5JL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2011.13824&json=true","fetch_graph":"https://pith.science/api/pith-number/SUEVI5JLHLNPELCPTAMSIQ7PWM/graph.json","fetch_events":"https://pith.science/api/pith-number/SUEVI5JLHLNPELCPTAMSIQ7PWM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM/action/storage_attestation","attest_author":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM/action/author_attestation","sign_citation":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM/action/citation_signature","submit_replication":"https://pith.science/pith/SUEVI5JLHLNPELCPTAMSIQ7PWM/action/replication_record"}},"created_at":"2026-07-05T02:23:19.779957+00:00","updated_at":"2026-07-05T02:23:19.779957+00:00"}