{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:IXSJ3BOKZHGCDH7KP7VGIYZSEL","short_pith_number":"pith:IXSJ3BOK","schema_version":"1.0","canonical_sha256":"45e49d85cac9cc219fea7fea64633222d43d104f2f2b48b72597d2863e588900","source":{"kind":"arxiv","id":"1709.06173","version":1},"attestation_state":"computed","paper":{"title":"Robustness of Neural Networks against Storage Media Errors","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.IT"],"primary_cat":"cs.IT","authors_text":"Chao Sun, Dejan Vucinic, Minghai Qin","submitted_at":"2017-09-18T21:37:55Z","abstract_excerpt":"We study the trade-offs between storage/bandwidth and prediction accuracy of neural networks that are stored in noisy media. Conventionally, it is assumed that all parameters (e.g., weight and biases) of a trained neural network are stored as binary arrays and are error-free. This assumption is based upon the implementation of error correction codes (ECCs) that correct potential bit flips in storage media. However, ECCs add storage overhead and cause bandwidth reduction when loading the trained parameters during the inference. We study the robustness of deep neural networks when bit errors exi"},"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":"1709.06173","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IT","submitted_at":"2017-09-18T21:37:55Z","cross_cats_sorted":["math.IT"],"title_canon_sha256":"8fc10ab754c9c27cb66c7c2daad073d04a7e2a582b11aa23bf4d21fc497f11d2","abstract_canon_sha256":"9ef4bcdadebe49abd41184d18cb7a8c4fbe2cbb143e4b0c86d82d0f556759ddc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:34:54.140802Z","signature_b64":"SiLzwPGZq0WhUg89TnxDi2QiJrgQxK/JYO8StBBvz+dudsuSwbUzP4nDtq+jU+HFlDhjxUAye+dflTtBhi38AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45e49d85cac9cc219fea7fea64633222d43d104f2f2b48b72597d2863e588900","last_reissued_at":"2026-05-18T00:34:54.140085Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:34:54.140085Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Robustness of Neural Networks against Storage Media Errors","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.IT"],"primary_cat":"cs.IT","authors_text":"Chao Sun, Dejan Vucinic, Minghai Qin","submitted_at":"2017-09-18T21:37:55Z","abstract_excerpt":"We study the trade-offs between storage/bandwidth and prediction accuracy of neural networks that are stored in noisy media. Conventionally, it is assumed that all parameters (e.g., weight and biases) of a trained neural network are stored as binary arrays and are error-free. This assumption is based upon the implementation of error correction codes (ECCs) that correct potential bit flips in storage media. However, ECCs add storage overhead and cause bandwidth reduction when loading the trained parameters during the inference. We study the robustness of deep neural networks when bit errors exi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1709.06173","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":""},"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":"1709.06173","created_at":"2026-05-18T00:34:54.140213+00:00"},{"alias_kind":"arxiv_version","alias_value":"1709.06173v1","created_at":"2026-05-18T00:34:54.140213+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1709.06173","created_at":"2026-05-18T00:34:54.140213+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.04563","citing_title":"RangeGuard: Efficient, Bounded Approximate Error Correction for Reliable DNNs","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07417","citing_title":"Effective and Memory-Efficient Alternatives to ECC for Reliable Large-Scale DNNs","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL","json":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL.json","graph_json":"https://pith.science/api/pith-number/IXSJ3BOKZHGCDH7KP7VGIYZSEL/graph.json","events_json":"https://pith.science/api/pith-number/IXSJ3BOKZHGCDH7KP7VGIYZSEL/events.json","paper":"https://pith.science/paper/IXSJ3BOK"},"agent_actions":{"view_html":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL","download_json":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL.json","view_paper":"https://pith.science/paper/IXSJ3BOK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1709.06173&json=true","fetch_graph":"https://pith.science/api/pith-number/IXSJ3BOKZHGCDH7KP7VGIYZSEL/graph.json","fetch_events":"https://pith.science/api/pith-number/IXSJ3BOKZHGCDH7KP7VGIYZSEL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL/action/storage_attestation","attest_author":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL/action/author_attestation","sign_citation":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL/action/citation_signature","submit_replication":"https://pith.science/pith/IXSJ3BOKZHGCDH7KP7VGIYZSEL/action/replication_record"}},"created_at":"2026-05-18T00:34:54.140213+00:00","updated_at":"2026-05-18T00:34:54.140213+00:00"}