{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2015:A7AQ4OXJ5GWDUF2URJ3XGABKQR","short_pith_number":"pith:A7AQ4OXJ","canonical_record":{"source":{"id":"1512.04906","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2015-12-15T19:29:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ccfa3a5b4a1112cf4c71ff50103858c5f8294aab5284864f00a5d95d5a7919ea","abstract_canon_sha256":"3ec1ce99d7b0bc89b41dde80901cdcf1c6141d723260a7cc98b99473bfbcea37"},"schema_version":"1.0"},"canonical_sha256":"07c10e3ae9e9ac3a17548a7773002a84790a359ce752337e4f9838b1f4601397","source":{"kind":"arxiv","id":"1512.04906","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1512.04906","created_at":"2026-05-18T01:24:15Z"},{"alias_kind":"arxiv_version","alias_value":"1512.04906v1","created_at":"2026-05-18T01:24:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1512.04906","created_at":"2026-05-18T01:24:15Z"},{"alias_kind":"pith_short_12","alias_value":"A7AQ4OXJ5GWD","created_at":"2026-05-18T12:29:10Z"},{"alias_kind":"pith_short_16","alias_value":"A7AQ4OXJ5GWDUF2U","created_at":"2026-05-18T12:29:10Z"},{"alias_kind":"pith_short_8","alias_value":"A7AQ4OXJ","created_at":"2026-05-18T12:29:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2015:A7AQ4OXJ5GWDUF2URJ3XGABKQR","target":"record","payload":{"canonical_record":{"source":{"id":"1512.04906","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2015-12-15T19:29:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"ccfa3a5b4a1112cf4c71ff50103858c5f8294aab5284864f00a5d95d5a7919ea","abstract_canon_sha256":"3ec1ce99d7b0bc89b41dde80901cdcf1c6141d723260a7cc98b99473bfbcea37"},"schema_version":"1.0"},"canonical_sha256":"07c10e3ae9e9ac3a17548a7773002a84790a359ce752337e4f9838b1f4601397","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:24:15.152908Z","signature_b64":"0mVVg9wp6IteqroeInPNQzyevHKE7hM74bl7gs8cSezpBzMUPkNd09nqJB/nfVR9uqEdta7SIZXbl1/25/2fCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"07c10e3ae9e9ac3a17548a7773002a84790a359ce752337e4f9838b1f4601397","last_reissued_at":"2026-05-18T01:24:15.152215Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:24:15.152215Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1512.04906","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T01:24:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6P5hnwEDlxwZyRifkx3ug4ua+cvh9/H63U8ZpEmKeP0IsbHae7KkbvvCUrxEzd7DZ8xAonsrkUkgK7DPpV/IDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T11:15:51.626224Z"},"content_sha256":"6b3be2d9b2b8ece2c29c46d10df292a3f91b139fffd815307c733c47f0bfe00e","schema_version":"1.0","event_id":"sha256:6b3be2d9b2b8ece2c29c46d10df292a3f91b139fffd815307c733c47f0bfe00e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2015:A7AQ4OXJ5GWDUF2URJ3XGABKQR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Strategies for Training Large Vocabulary Neural Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"David Grangier, Michael Auli, Welin Chen","submitted_at":"2015-12-15T19:29:01Z","abstract_excerpt":"Training neural network language models over large vocabularies is still computationally very costly compared to count-based models such as Kneser-Ney. At the same time, neural language models are gaining popularity for many applications such as speech recognition and machine translation whose success depends on scalability. We present a systematic comparison of strategies to represent and train large vocabularies, including softmax, hierarchical softmax, target sampling, noise contrastive estimation and self normalization. We further extend self normalization to be a proper estimator of likel"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1512.04906","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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T01:24:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/jSShpfxmbjUqgLtGH1XVQoVM4Xh18u/S+ovi1tmZoyNmoOIqIvZK8ZBKos2MbRbHf2W9a+BWCFAggcaOOXcDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T11:15:51.626575Z"},"content_sha256":"f42442c793e8ba4ffc529f5014bdc9747add72576444acc260730ed904769a80","schema_version":"1.0","event_id":"sha256:f42442c793e8ba4ffc529f5014bdc9747add72576444acc260730ed904769a80"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR/bundle.json","state_url":"https://pith.science/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR/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":"2026-06-09T11:15:51Z","links":{"resolver":"https://pith.science/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR","bundle":"https://pith.science/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR/bundle.json","state":"https://pith.science/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/A7AQ4OXJ5GWDUF2URJ3XGABKQR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2015:A7AQ4OXJ5GWDUF2URJ3XGABKQR","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":"3ec1ce99d7b0bc89b41dde80901cdcf1c6141d723260a7cc98b99473bfbcea37","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2015-12-15T19:29:01Z","title_canon_sha256":"ccfa3a5b4a1112cf4c71ff50103858c5f8294aab5284864f00a5d95d5a7919ea"},"schema_version":"1.0","source":{"id":"1512.04906","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1512.04906","created_at":"2026-05-18T01:24:15Z"},{"alias_kind":"arxiv_version","alias_value":"1512.04906v1","created_at":"2026-05-18T01:24:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1512.04906","created_at":"2026-05-18T01:24:15Z"},{"alias_kind":"pith_short_12","alias_value":"A7AQ4OXJ5GWD","created_at":"2026-05-18T12:29:10Z"},{"alias_kind":"pith_short_16","alias_value":"A7AQ4OXJ5GWDUF2U","created_at":"2026-05-18T12:29:10Z"},{"alias_kind":"pith_short_8","alias_value":"A7AQ4OXJ","created_at":"2026-05-18T12:29:10Z"}],"graph_snapshots":[{"event_id":"sha256:f42442c793e8ba4ffc529f5014bdc9747add72576444acc260730ed904769a80","target":"graph","created_at":"2026-05-18T01:24:15Z","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":"Training neural network language models over large vocabularies is still computationally very costly compared to count-based models such as Kneser-Ney. At the same time, neural language models are gaining popularity for many applications such as speech recognition and machine translation whose success depends on scalability. We present a systematic comparison of strategies to represent and train large vocabularies, including softmax, hierarchical softmax, target sampling, noise contrastive estimation and self normalization. We further extend self normalization to be a proper estimator of likel","authors_text":"David Grangier, Michael Auli, Welin Chen","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2015-12-15T19:29:01Z","title":"Strategies for Training Large Vocabulary Neural Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1512.04906","kind":"arxiv","version":1},"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":"sha256:6b3be2d9b2b8ece2c29c46d10df292a3f91b139fffd815307c733c47f0bfe00e","target":"record","created_at":"2026-05-18T01:24:15Z","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":"3ec1ce99d7b0bc89b41dde80901cdcf1c6141d723260a7cc98b99473bfbcea37","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2015-12-15T19:29:01Z","title_canon_sha256":"ccfa3a5b4a1112cf4c71ff50103858c5f8294aab5284864f00a5d95d5a7919ea"},"schema_version":"1.0","source":{"id":"1512.04906","kind":"arxiv","version":1}},"canonical_sha256":"07c10e3ae9e9ac3a17548a7773002a84790a359ce752337e4f9838b1f4601397","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"07c10e3ae9e9ac3a17548a7773002a84790a359ce752337e4f9838b1f4601397","first_computed_at":"2026-05-18T01:24:15.152215Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T01:24:15.152215Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0mVVg9wp6IteqroeInPNQzyevHKE7hM74bl7gs8cSezpBzMUPkNd09nqJB/nfVR9uqEdta7SIZXbl1/25/2fCg==","signature_status":"signed_v1","signed_at":"2026-05-18T01:24:15.152908Z","signed_message":"canonical_sha256_bytes"},"source_id":"1512.04906","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6b3be2d9b2b8ece2c29c46d10df292a3f91b139fffd815307c733c47f0bfe00e","sha256:f42442c793e8ba4ffc529f5014bdc9747add72576444acc260730ed904769a80"],"state_sha256":"231d8bf2a16c960907443237ba972de066e91b57b3001f69fcc9b5f603d8cf0b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"C86YwSVwt48+Vmn23skqQYhPs2PYUSmBykyIxa6TYAxcU+aj7D0D/KkTZXg32DaNip4vo2wV5bDoTm46h+K4Cw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-09T11:15:51.628331Z","bundle_sha256":"44a7f607e2f6dcfb0b211c11956fe4af4f6fa2675cefc3a824bf01f781d4c33c"}}