{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:VDA6XFVVOZAAASTMHQZXCO6DY3","short_pith_number":"pith:VDA6XFVV","schema_version":"1.0","canonical_sha256":"a8c1eb96b57640004a6c3c33713bc3c6e6667a7a3ccb13601beb309ae5f2bb36","source":{"kind":"arxiv","id":"1810.05486","version":1},"attestation_state":"computed","paper":{"title":"Training Deep Neural Network in Limited Precision","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.NE","authors_text":"Hyunsun Park, Jun Haeng Lee, Sangwon Ha, Seungwon Lee, Youngmin Oh","submitted_at":"2018-10-12T12:58:18Z","abstract_excerpt":"Energy and resource efficient training of DNNs will greatly extend the applications of deep learning. However, there are three major obstacles which mandate accurate calculation in high precision. In this paper, we tackle two of them related to the loss of gradients during parameter update and backpropagation through a softmax nonlinearity layer in low precision training. We implemented SGD with Kahan summation by employing an additional parameter to virtually extend the bit-width of the parameters for a reliable parameter update. We also proposed a simple guideline to help select the appropri"},"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":"1810.05486","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2018-10-12T12:58:18Z","cross_cats_sorted":[],"title_canon_sha256":"6b019134d259721f9264e75484e56f452a705d2a36a3cad2aaadcf5570152b41","abstract_canon_sha256":"9c9db3b38f7354c4753ac0caa37e8d5b4bbb8489893380c6638d382741a77f8e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:03:30.371902Z","signature_b64":"buDoYf17ygjGttIP8hemXXEzas5mB65BT/FrqcHSyjTxwwU+/jGX0EmMSGQiI5XI3DBlkFLTHJE1CnJ6By0RDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a8c1eb96b57640004a6c3c33713bc3c6e6667a7a3ccb13601beb309ae5f2bb36","last_reissued_at":"2026-05-18T00:03:30.371207Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:03:30.371207Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training Deep Neural Network in Limited Precision","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.NE","authors_text":"Hyunsun Park, Jun Haeng Lee, Sangwon Ha, Seungwon Lee, Youngmin Oh","submitted_at":"2018-10-12T12:58:18Z","abstract_excerpt":"Energy and resource efficient training of DNNs will greatly extend the applications of deep learning. However, there are three major obstacles which mandate accurate calculation in high precision. In this paper, we tackle two of them related to the loss of gradients during parameter update and backpropagation through a softmax nonlinearity layer in low precision training. We implemented SGD with Kahan summation by employing an additional parameter to virtually extend the bit-width of the parameters for a reliable parameter update. We also proposed a simple guideline to help select the appropri"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1810.05486","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":"1810.05486","created_at":"2026-05-18T00:03:30.371324+00:00"},{"alias_kind":"arxiv_version","alias_value":"1810.05486v1","created_at":"2026-05-18T00:03:30.371324+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1810.05486","created_at":"2026-05-18T00:03:30.371324+00:00"},{"alias_kind":"pith_short_12","alias_value":"VDA6XFVVOZAA","created_at":"2026-05-18T12:32:59.047623+00:00"},{"alias_kind":"pith_short_16","alias_value":"VDA6XFVVOZAAASTM","created_at":"2026-05-18T12:32:59.047623+00:00"},{"alias_kind":"pith_short_8","alias_value":"VDA6XFVV","created_at":"2026-05-18T12:32:59.047623+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/VDA6XFVVOZAAASTMHQZXCO6DY3","json":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3.json","graph_json":"https://pith.science/api/pith-number/VDA6XFVVOZAAASTMHQZXCO6DY3/graph.json","events_json":"https://pith.science/api/pith-number/VDA6XFVVOZAAASTMHQZXCO6DY3/events.json","paper":"https://pith.science/paper/VDA6XFVV"},"agent_actions":{"view_html":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3","download_json":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3.json","view_paper":"https://pith.science/paper/VDA6XFVV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1810.05486&json=true","fetch_graph":"https://pith.science/api/pith-number/VDA6XFVVOZAAASTMHQZXCO6DY3/graph.json","fetch_events":"https://pith.science/api/pith-number/VDA6XFVVOZAAASTMHQZXCO6DY3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3/action/storage_attestation","attest_author":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3/action/author_attestation","sign_citation":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3/action/citation_signature","submit_replication":"https://pith.science/pith/VDA6XFVVOZAAASTMHQZXCO6DY3/action/replication_record"}},"created_at":"2026-05-18T00:03:30.371324+00:00","updated_at":"2026-05-18T00:03:30.371324+00:00"}