{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:UJMPURX5V5EICUCJQNZLYKKHTQ","short_pith_number":"pith:UJMPURX5","schema_version":"1.0","canonical_sha256":"a258fa46fdaf488150498372bc29479c159e90ade967af9f7ee747930ac47411","source":{"kind":"arxiv","id":"1910.06909","version":2},"attestation_state":"computed","paper":{"title":"OverQ: Opportunistic Outlier Quantization for Neural Network Accelerators","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Christopher De Sa, Jordan Dotzel, Preslav Ivanov, Ritchie Zhao, Zhanqiu Hu, Zhiru Zhang","submitted_at":"2019-10-13T23:21:59Z","abstract_excerpt":"Outliers in weights and activations pose a key challenge for fixed-point quantization of neural networks. While they can be addressed by fine-tuning, this is not practical for ML service providers (e.g., Google or Microsoft) who often receive customer models without training data. Specialized hardware for handling activation outliers can enable low-precision neural networks, but at the cost of nontrivial area overhead. We instead propose overwrite quantization (OverQ), a lightweight hardware technique that opportunistically increases bitwidth for activation outliers by overwriting nearby zeros"},"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":"1910.06909","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-10-13T23:21:59Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"0b98e7b5021a5fac631404fc297a0980ab7a015f0294c5c4b40741434b09c91f","abstract_canon_sha256":"8a42b89319e3a9a450ab6437bfcbfd8cf2d92a35e52170d515054b1e5242e830"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:43:50.845843Z","signature_b64":"1/Too+c4gtzJprGgEGV6a66loamovxVB2Q5wE6h6G43zjpeNTWelOeDA8i+ywoqtyc6KN2hoozUPg+mSJG6SDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a258fa46fdaf488150498372bc29479c159e90ade967af9f7ee747930ac47411","last_reissued_at":"2026-07-05T02:43:50.845383Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:43:50.845383Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OverQ: Opportunistic Outlier Quantization for Neural Network Accelerators","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Christopher De Sa, Jordan Dotzel, Preslav Ivanov, Ritchie Zhao, Zhanqiu Hu, Zhiru Zhang","submitted_at":"2019-10-13T23:21:59Z","abstract_excerpt":"Outliers in weights and activations pose a key challenge for fixed-point quantization of neural networks. While they can be addressed by fine-tuning, this is not practical for ML service providers (e.g., Google or Microsoft) who often receive customer models without training data. Specialized hardware for handling activation outliers can enable low-precision neural networks, but at the cost of nontrivial area overhead. We instead propose overwrite quantization (OverQ), a lightweight hardware technique that opportunistically increases bitwidth for activation outliers by overwriting nearby zeros"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.06909","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/1910.06909/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":"1910.06909","created_at":"2026-07-05T02:43:50.845447+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.06909v2","created_at":"2026-07-05T02:43:50.845447+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.06909","created_at":"2026-07-05T02:43:50.845447+00:00"},{"alias_kind":"pith_short_12","alias_value":"UJMPURX5V5EI","created_at":"2026-07-05T02:43:50.845447+00:00"},{"alias_kind":"pith_short_16","alias_value":"UJMPURX5V5EICUCJ","created_at":"2026-07-05T02:43:50.845447+00:00"},{"alias_kind":"pith_short_8","alias_value":"UJMPURX5","created_at":"2026-07-05T02:43:50.845447+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ","json":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ.json","graph_json":"https://pith.science/api/pith-number/UJMPURX5V5EICUCJQNZLYKKHTQ/graph.json","events_json":"https://pith.science/api/pith-number/UJMPURX5V5EICUCJQNZLYKKHTQ/events.json","paper":"https://pith.science/paper/UJMPURX5"},"agent_actions":{"view_html":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ","download_json":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ.json","view_paper":"https://pith.science/paper/UJMPURX5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.06909&json=true","fetch_graph":"https://pith.science/api/pith-number/UJMPURX5V5EICUCJQNZLYKKHTQ/graph.json","fetch_events":"https://pith.science/api/pith-number/UJMPURX5V5EICUCJQNZLYKKHTQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ/action/storage_attestation","attest_author":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ/action/author_attestation","sign_citation":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ/action/citation_signature","submit_replication":"https://pith.science/pith/UJMPURX5V5EICUCJQNZLYKKHTQ/action/replication_record"}},"created_at":"2026-07-05T02:43:50.845447+00:00","updated_at":"2026-07-05T02:43:50.845447+00:00"}