{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:ACOHDHWEAOJLFFBYXHPNNS3NDE","short_pith_number":"pith:ACOHDHWE","canonical_record":{"source":{"id":"1808.04444","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-09T18:44:38Z","cross_cats_sorted":["cs.AI","cs.LG","stat.ML"],"title_canon_sha256":"223c492c6cb0e0e1af30c360e52c22884c14521a00716ba00c60a0ca9271f543","abstract_canon_sha256":"244e384bc2171ebd2e9dae3bae885e15ac0f77e2edaaeb743e4fac84a9d66f90"},"schema_version":"1.0"},"canonical_sha256":"009c719ec40392b29438b9ded6cb6d193b20f988aa96f7412afcf856f6a58a57","source":{"kind":"arxiv","id":"1808.04444","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1808.04444","created_at":"2026-05-17T23:58:47Z"},{"alias_kind":"arxiv_version","alias_value":"1808.04444v2","created_at":"2026-05-17T23:58:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.04444","created_at":"2026-05-17T23:58:47Z"},{"alias_kind":"pith_short_12","alias_value":"ACOHDHWEAOJL","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_16","alias_value":"ACOHDHWEAOJLFFBY","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_8","alias_value":"ACOHDHWE","created_at":"2026-05-18T12:32:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:ACOHDHWEAOJLFFBYXHPNNS3NDE","target":"record","payload":{"canonical_record":{"source":{"id":"1808.04444","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-09T18:44:38Z","cross_cats_sorted":["cs.AI","cs.LG","stat.ML"],"title_canon_sha256":"223c492c6cb0e0e1af30c360e52c22884c14521a00716ba00c60a0ca9271f543","abstract_canon_sha256":"244e384bc2171ebd2e9dae3bae885e15ac0f77e2edaaeb743e4fac84a9d66f90"},"schema_version":"1.0"},"canonical_sha256":"009c719ec40392b29438b9ded6cb6d193b20f988aa96f7412afcf856f6a58a57","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:58:47.692660Z","signature_b64":"eXy+wFFtM07v/jyDCdkJ+8lFRTzOIWEf4MTU0PfM8VhLwW5CxAVhZXV5Oq4jARYJlDKkGhdj1+5NFS5DuuP5AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"009c719ec40392b29438b9ded6cb6d193b20f988aa96f7412afcf856f6a58a57","last_reissued_at":"2026-05-17T23:58:47.692179Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:58:47.692179Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1808.04444","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:58:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gohTYV4g/kcCUgAoSnPDvXVOh5zn+g4rBkawmgW3PwKGl5P8yjz0QTdQdsWyj4r+o7bZmhfWvb4jobdFmzlACQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-10T10:16:23.943186Z"},"content_sha256":"f6a32d709455c8e79b5481744a66867aa9ee451e7975809f9e699a9194db72cf","schema_version":"1.0","event_id":"sha256:f6a32d709455c8e79b5481744a66867aa9ee451e7975809f9e699a9194db72cf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:ACOHDHWEAOJLFFBYXHPNNS3NDE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Character-Level Language Modeling with Deeper Self-Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Dokook Choe, Llion Jones, Mandy Guo, Noah Constant, Rami Al-Rfou","submitted_at":"2018-08-09T18:44:38Z","abstract_excerpt":"LSTMs and other RNN variants have shown strong performance on character-level language modeling. These models are typically trained using truncated backpropagation through time, and it is common to assume that their success stems from their ability to remember long-term contexts. In this paper, we show that a deep (64-layer) transformer model with fixed context outperforms RNN variants by a large margin, achieving state of the art on two popular benchmarks: 1.13 bits per character on text8 and 1.06 on enwik8. To get good results at this depth, we show that it is important to add auxiliary loss"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.04444","kind":"arxiv","version":2},"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":"2026-05-17T23:58:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eO/pKenWIfTFeizBOf8oMPF+uxo826ugOW52Emq4jaYzo7kG8UvuN66mVV7nY2RSyBcSgQ0+1Li08yK+dQ8CBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-10T10:16:23.943917Z"},"content_sha256":"dde801555079b918f7aa655b3d3a8c8559aca0ded2b4778e1d4183b36ad7a3cd","schema_version":"1.0","event_id":"sha256:dde801555079b918f7aa655b3d3a8c8559aca0ded2b4778e1d4183b36ad7a3cd"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE/bundle.json","state_url":"https://pith.science/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE/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":"2026-06-10T10:16:23Z","links":{"resolver":"https://pith.science/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE","bundle":"https://pith.science/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE/bundle.json","state":"https://pith.science/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ACOHDHWEAOJLFFBYXHPNNS3NDE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:ACOHDHWEAOJLFFBYXHPNNS3NDE","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":"244e384bc2171ebd2e9dae3bae885e15ac0f77e2edaaeb743e4fac84a9d66f90","cross_cats_sorted":["cs.AI","cs.LG","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-09T18:44:38Z","title_canon_sha256":"223c492c6cb0e0e1af30c360e52c22884c14521a00716ba00c60a0ca9271f543"},"schema_version":"1.0","source":{"id":"1808.04444","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1808.04444","created_at":"2026-05-17T23:58:47Z"},{"alias_kind":"arxiv_version","alias_value":"1808.04444v2","created_at":"2026-05-17T23:58:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.04444","created_at":"2026-05-17T23:58:47Z"},{"alias_kind":"pith_short_12","alias_value":"ACOHDHWEAOJL","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_16","alias_value":"ACOHDHWEAOJLFFBY","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_8","alias_value":"ACOHDHWE","created_at":"2026-05-18T12:32:13Z"}],"graph_snapshots":[{"event_id":"sha256:dde801555079b918f7aa655b3d3a8c8559aca0ded2b4778e1d4183b36ad7a3cd","target":"graph","created_at":"2026-05-17T23:58:47Z","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":"LSTMs and other RNN variants have shown strong performance on character-level language modeling. These models are typically trained using truncated backpropagation through time, and it is common to assume that their success stems from their ability to remember long-term contexts. In this paper, we show that a deep (64-layer) transformer model with fixed context outperforms RNN variants by a large margin, achieving state of the art on two popular benchmarks: 1.13 bits per character on text8 and 1.06 on enwik8. To get good results at this depth, we show that it is important to add auxiliary loss","authors_text":"Dokook Choe, Llion Jones, Mandy Guo, Noah Constant, Rami Al-Rfou","cross_cats":["cs.AI","cs.LG","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-09T18:44:38Z","title":"Character-Level Language Modeling with Deeper Self-Attention"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.04444","kind":"arxiv","version":2},"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:f6a32d709455c8e79b5481744a66867aa9ee451e7975809f9e699a9194db72cf","target":"record","created_at":"2026-05-17T23:58:47Z","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":"244e384bc2171ebd2e9dae3bae885e15ac0f77e2edaaeb743e4fac84a9d66f90","cross_cats_sorted":["cs.AI","cs.LG","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-09T18:44:38Z","title_canon_sha256":"223c492c6cb0e0e1af30c360e52c22884c14521a00716ba00c60a0ca9271f543"},"schema_version":"1.0","source":{"id":"1808.04444","kind":"arxiv","version":2}},"canonical_sha256":"009c719ec40392b29438b9ded6cb6d193b20f988aa96f7412afcf856f6a58a57","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"009c719ec40392b29438b9ded6cb6d193b20f988aa96f7412afcf856f6a58a57","first_computed_at":"2026-05-17T23:58:47.692179Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:58:47.692179Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eXy+wFFtM07v/jyDCdkJ+8lFRTzOIWEf4MTU0PfM8VhLwW5CxAVhZXV5Oq4jARYJlDKkGhdj1+5NFS5DuuP5AA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:58:47.692660Z","signed_message":"canonical_sha256_bytes"},"source_id":"1808.04444","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f6a32d709455c8e79b5481744a66867aa9ee451e7975809f9e699a9194db72cf","sha256:dde801555079b918f7aa655b3d3a8c8559aca0ded2b4778e1d4183b36ad7a3cd"],"state_sha256":"6057fa692296bdb25992c42b4f2ed8e610150cc185fdcca4ee91d7619f9b8a31"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3asjPMxKvhB3DdItQaO7iv5XnNRABJ6e3KrwQQtdQHqxiDHnkFAVYSd0oQQmUPiDkL0nP0hgeXHvtXDNf/b9Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-10T10:16:23.948438Z","bundle_sha256":"ed7dc5c29f3a204bcd0ce0d7284fc1522f7c8eabf1a6a3a47097bcf7f86a89e3"}}