{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:GI22VBHIDQ3NDYRVJGLS7EVEXF","short_pith_number":"pith:GI22VBHI","schema_version":"1.0","canonical_sha256":"3235aa84e81c36d1e23549972f92a4b941bc1cf446b3dcd0de88b086efaf1896","source":{"kind":"arxiv","id":"1804.01526","version":4},"attestation_state":"computed","paper":{"title":"Training DNNs with Hybrid Block Floating Point","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Babak Falsafi, Mario Drumond, Martin Jaggi, Tao Lin","submitted_at":"2018-04-04T09:40:59Z","abstract_excerpt":"The wide adoption of DNNs has given birth to unrelenting computing requirements, forcing datacenter operators to adopt domain-specific accelerators to train them. These accelerators typically employ densely packed full precision floating-point arithmetic to maximize performance per area. Ongoing research efforts seek to further increase that performance density by replacing floating-point with fixed-point arithmetic. However, a significant roadblock for these attempts has been fixed point's narrow dynamic range, which is insufficient for DNN training convergence. We identify block floating poi"},"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":"1804.01526","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-04-04T09:40:59Z","cross_cats_sorted":["cs.NA","stat.ML"],"title_canon_sha256":"7d5e2d53dcbc56761ffacd0ce82fcfed7057b541ab8f9e3b7b02e147351458b2","abstract_canon_sha256":"13ee8fb0a6f6f4cfa8aee9e22f2ff781f45783d6ad52732614ec614a807f7275"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:59:22.832736Z","signature_b64":"HNp9Y3ZsaxUNgrPNwyyUs3x0WMWJLEwD7dT/6qs965QRFJJYYP/zcehp0NYltB7/JEkAV9tihSVh8VCyVjx9AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3235aa84e81c36d1e23549972f92a4b941bc1cf446b3dcd0de88b086efaf1896","last_reissued_at":"2026-05-17T23:59:22.832401Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:59:22.832401Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training DNNs with Hybrid Block Floating Point","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Babak Falsafi, Mario Drumond, Martin Jaggi, Tao Lin","submitted_at":"2018-04-04T09:40:59Z","abstract_excerpt":"The wide adoption of DNNs has given birth to unrelenting computing requirements, forcing datacenter operators to adopt domain-specific accelerators to train them. These accelerators typically employ densely packed full precision floating-point arithmetic to maximize performance per area. Ongoing research efforts seek to further increase that performance density by replacing floating-point with fixed-point arithmetic. However, a significant roadblock for these attempts has been fixed point's narrow dynamic range, which is insufficient for DNN training convergence. We identify block floating poi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1804.01526","kind":"arxiv","version":4},"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":"1804.01526","created_at":"2026-05-17T23:59:22.832455+00:00"},{"alias_kind":"arxiv_version","alias_value":"1804.01526v4","created_at":"2026-05-17T23:59:22.832455+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1804.01526","created_at":"2026-05-17T23:59:22.832455+00:00"},{"alias_kind":"pith_short_12","alias_value":"GI22VBHIDQ3N","created_at":"2026-05-18T12:32:25.280505+00:00"},{"alias_kind":"pith_short_16","alias_value":"GI22VBHIDQ3NDYRV","created_at":"2026-05-18T12:32:25.280505+00:00"},{"alias_kind":"pith_short_8","alias_value":"GI22VBHI","created_at":"2026-05-18T12:32:25.280505+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/GI22VBHIDQ3NDYRVJGLS7EVEXF","json":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF.json","graph_json":"https://pith.science/api/pith-number/GI22VBHIDQ3NDYRVJGLS7EVEXF/graph.json","events_json":"https://pith.science/api/pith-number/GI22VBHIDQ3NDYRVJGLS7EVEXF/events.json","paper":"https://pith.science/paper/GI22VBHI"},"agent_actions":{"view_html":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF","download_json":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF.json","view_paper":"https://pith.science/paper/GI22VBHI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1804.01526&json=true","fetch_graph":"https://pith.science/api/pith-number/GI22VBHIDQ3NDYRVJGLS7EVEXF/graph.json","fetch_events":"https://pith.science/api/pith-number/GI22VBHIDQ3NDYRVJGLS7EVEXF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF/action/storage_attestation","attest_author":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF/action/author_attestation","sign_citation":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF/action/citation_signature","submit_replication":"https://pith.science/pith/GI22VBHIDQ3NDYRVJGLS7EVEXF/action/replication_record"}},"created_at":"2026-05-17T23:59:22.832455+00:00","updated_at":"2026-05-17T23:59:22.832455+00:00"}