{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2014:NCLOTRGSQ472LOUTR5V2UO5T7X","short_pith_number":"pith:NCLOTRGS","canonical_record":{"source":{"id":"1412.7091","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2014-12-22T18:51:08Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"a183fcca645a3bdb87086317e84dbf778b1730bb16e1e30f19ed03a24a71b7bd","abstract_canon_sha256":"5c9f664129b8d2aa5db568b4830681fb056c2eea79eeaa98ef8cf1a828f47f8e"},"schema_version":"1.0"},"canonical_sha256":"6896e9c4d2873fa5ba938f6baa3bb3fdc99d0c9afae9d11679c685a5c835ce7f","source":{"kind":"arxiv","id":"1412.7091","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1412.7091","created_at":"2026-05-18T01:36:58Z"},{"alias_kind":"arxiv_version","alias_value":"1412.7091v3","created_at":"2026-05-18T01:36:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1412.7091","created_at":"2026-05-18T01:36:58Z"},{"alias_kind":"pith_short_12","alias_value":"NCLOTRGSQ472","created_at":"2026-05-18T12:28:41Z"},{"alias_kind":"pith_short_16","alias_value":"NCLOTRGSQ472LOUT","created_at":"2026-05-18T12:28:41Z"},{"alias_kind":"pith_short_8","alias_value":"NCLOTRGS","created_at":"2026-05-18T12:28:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2014:NCLOTRGSQ472LOUTR5V2UO5T7X","target":"record","payload":{"canonical_record":{"source":{"id":"1412.7091","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2014-12-22T18:51:08Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"a183fcca645a3bdb87086317e84dbf778b1730bb16e1e30f19ed03a24a71b7bd","abstract_canon_sha256":"5c9f664129b8d2aa5db568b4830681fb056c2eea79eeaa98ef8cf1a828f47f8e"},"schema_version":"1.0"},"canonical_sha256":"6896e9c4d2873fa5ba938f6baa3bb3fdc99d0c9afae9d11679c685a5c835ce7f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:36:58.165650Z","signature_b64":"0O4wrHFiCWBE0rAX5DtNhow3i4wXI23J4LBe7QhTn0ER2GS4+BgICKKDNT7RIceY3bbxIsgodXMy080eR5hWDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6896e9c4d2873fa5ba938f6baa3bb3fdc99d0c9afae9d11679c685a5c835ce7f","last_reissued_at":"2026-05-18T01:36:58.165026Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:36:58.165026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1412.7091","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T01:36:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"o/i/38hmXowkiqqVLLvDD2EjsZInNEVYT/tjpc5BsGuOMBjrOeywroWb23gmAJlcqIZOUT/91WX6XTb+c1JJCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T00:44:19.354905Z"},"content_sha256":"a9a3e1a0a7517055365da6ba82819c5180e56d70cd90e1accb063bbc5a7dd46d","schema_version":"1.0","event_id":"sha256:a9a3e1a0a7517055365da6ba82819c5180e56d70cd90e1accb063bbc5a7dd46d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2014:NCLOTRGSQ472LOUTR5V2UO5T7X","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Efficient Exact Gradient Update for training Deep Networks with Very Large Sparse Targets","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.NE","authors_text":"Alexandre de Br\\'ebisson, Pascal Vincent, Xavier Bouthillier","submitted_at":"2014-12-22T18:51:08Z","abstract_excerpt":"An important class of problems involves training deep neural networks with sparse prediction targets of very high dimension D. These occur naturally in e.g. neural language models or the learning of word-embeddings, often posed as predicting the probability of next words among a vocabulary of size D (e.g. 200 000). Computing the equally large, but typically non-sparse D-dimensional output vector from a last hidden layer of reasonable dimension d (e.g. 500) incurs a prohibitive O(Dd) computational cost for each example, as does updating the D x d output weight matrix and computing the gradient "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1412.7091","kind":"arxiv","version":3},"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T01:36:58Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5ZQbyjNt9dYmK+BolubEhznE4ditoUuoJiMfwsAwqLng/hp897rY/DyD1SL6Kp0/LbDgrSYBmrR7Oh3TYNuhDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T00:44:19.355541Z"},"content_sha256":"121e3570a8b21eb866c69e010c20067e05418ed802a868050d81a711813a5038","schema_version":"1.0","event_id":"sha256:121e3570a8b21eb866c69e010c20067e05418ed802a868050d81a711813a5038"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NCLOTRGSQ472LOUTR5V2UO5T7X/bundle.json","state_url":"https://pith.science/pith/NCLOTRGSQ472LOUTR5V2UO5T7X/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NCLOTRGSQ472LOUTR5V2UO5T7X/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-26T00:44:19Z","links":{"resolver":"https://pith.science/pith/NCLOTRGSQ472LOUTR5V2UO5T7X","bundle":"https://pith.science/pith/NCLOTRGSQ472LOUTR5V2UO5T7X/bundle.json","state":"https://pith.science/pith/NCLOTRGSQ472LOUTR5V2UO5T7X/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NCLOTRGSQ472LOUTR5V2UO5T7X/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2014:NCLOTRGSQ472LOUTR5V2UO5T7X","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"5c9f664129b8d2aa5db568b4830681fb056c2eea79eeaa98ef8cf1a828f47f8e","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2014-12-22T18:51:08Z","title_canon_sha256":"a183fcca645a3bdb87086317e84dbf778b1730bb16e1e30f19ed03a24a71b7bd"},"schema_version":"1.0","source":{"id":"1412.7091","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1412.7091","created_at":"2026-05-18T01:36:58Z"},{"alias_kind":"arxiv_version","alias_value":"1412.7091v3","created_at":"2026-05-18T01:36:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1412.7091","created_at":"2026-05-18T01:36:58Z"},{"alias_kind":"pith_short_12","alias_value":"NCLOTRGSQ472","created_at":"2026-05-18T12:28:41Z"},{"alias_kind":"pith_short_16","alias_value":"NCLOTRGSQ472LOUT","created_at":"2026-05-18T12:28:41Z"},{"alias_kind":"pith_short_8","alias_value":"NCLOTRGS","created_at":"2026-05-18T12:28:41Z"}],"graph_snapshots":[{"event_id":"sha256:121e3570a8b21eb866c69e010c20067e05418ed802a868050d81a711813a5038","target":"graph","created_at":"2026-05-18T01:36:58Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"An important class of problems involves training deep neural networks with sparse prediction targets of very high dimension D. These occur naturally in e.g. neural language models or the learning of word-embeddings, often posed as predicting the probability of next words among a vocabulary of size D (e.g. 200 000). Computing the equally large, but typically non-sparse D-dimensional output vector from a last hidden layer of reasonable dimension d (e.g. 500) incurs a prohibitive O(Dd) computational cost for each example, as does updating the D x d output weight matrix and computing the gradient ","authors_text":"Alexandre de Br\\'ebisson, Pascal Vincent, Xavier Bouthillier","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2014-12-22T18:51:08Z","title":"Efficient Exact Gradient Update for training Deep Networks with Very Large Sparse Targets"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1412.7091","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a9a3e1a0a7517055365da6ba82819c5180e56d70cd90e1accb063bbc5a7dd46d","target":"record","created_at":"2026-05-18T01:36:58Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"5c9f664129b8d2aa5db568b4830681fb056c2eea79eeaa98ef8cf1a828f47f8e","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.NE","submitted_at":"2014-12-22T18:51:08Z","title_canon_sha256":"a183fcca645a3bdb87086317e84dbf778b1730bb16e1e30f19ed03a24a71b7bd"},"schema_version":"1.0","source":{"id":"1412.7091","kind":"arxiv","version":3}},"canonical_sha256":"6896e9c4d2873fa5ba938f6baa3bb3fdc99d0c9afae9d11679c685a5c835ce7f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6896e9c4d2873fa5ba938f6baa3bb3fdc99d0c9afae9d11679c685a5c835ce7f","first_computed_at":"2026-05-18T01:36:58.165026Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T01:36:58.165026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0O4wrHFiCWBE0rAX5DtNhow3i4wXI23J4LBe7QhTn0ER2GS4+BgICKKDNT7RIceY3bbxIsgodXMy080eR5hWDQ==","signature_status":"signed_v1","signed_at":"2026-05-18T01:36:58.165650Z","signed_message":"canonical_sha256_bytes"},"source_id":"1412.7091","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a9a3e1a0a7517055365da6ba82819c5180e56d70cd90e1accb063bbc5a7dd46d","sha256:121e3570a8b21eb866c69e010c20067e05418ed802a868050d81a711813a5038"],"state_sha256":"99823847893561d8af6e3e1bfb1d31e09a11cc04f82f346e043a7d1dbf2bca38"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LrdgcpIXgMMWk6oGY3LQZJs5gGTtuN1Ft2eknu3agTL6LmyyaqsCpK2RcFW8ZbgOHYiR9uSpTQkqkXXyGBEYBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T00:44:19.358855Z","bundle_sha256":"e04944b226a2c471021a0f30f27489b94b8842a1318ced5b5d1fd142ede67894"}}