{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:QPVLLC37DCCC5MTP56I2FUUI5V","short_pith_number":"pith:QPVLLC37","canonical_record":{"source":{"id":"1805.05062","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-05-14T08:37:50Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"a0fc414edb0769912df155183fb4f68dd2d6c8fb1b6ea331ebca94437832420f","abstract_canon_sha256":"d7c5b89134b17ddaeba2ea71e0870cae74f6cac89be524aff865241b1db2cf1c"},"schema_version":"1.0"},"canonical_sha256":"83eab58b7f18842eb26fef91a2d288ed4e2e2d1ad115740fd9326779b42658ca","source":{"kind":"arxiv","id":"1805.05062","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1805.05062","created_at":"2026-05-18T00:16:03Z"},{"alias_kind":"arxiv_version","alias_value":"1805.05062v1","created_at":"2026-05-18T00:16:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1805.05062","created_at":"2026-05-18T00:16:03Z"},{"alias_kind":"pith_short_12","alias_value":"QPVLLC37DCCC","created_at":"2026-05-18T12:32:46Z"},{"alias_kind":"pith_short_16","alias_value":"QPVLLC37DCCC5MTP","created_at":"2026-05-18T12:32:46Z"},{"alias_kind":"pith_short_8","alias_value":"QPVLLC37","created_at":"2026-05-18T12:32:46Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:QPVLLC37DCCC5MTP56I2FUUI5V","target":"record","payload":{"canonical_record":{"source":{"id":"1805.05062","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-05-14T08:37:50Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"a0fc414edb0769912df155183fb4f68dd2d6c8fb1b6ea331ebca94437832420f","abstract_canon_sha256":"d7c5b89134b17ddaeba2ea71e0870cae74f6cac89be524aff865241b1db2cf1c"},"schema_version":"1.0"},"canonical_sha256":"83eab58b7f18842eb26fef91a2d288ed4e2e2d1ad115740fd9326779b42658ca","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:16:03.960018Z","signature_b64":"Gzan9tQKj0vrJ2YvyPbaUyQQ9CI1D9lsd8vhfKLUsSu9RKYqzrhPxSKjknc5gkp+7Fys1jC/D+42gBS3WwSLAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"83eab58b7f18842eb26fef91a2d288ed4e2e2d1ad115740fd9326779b42658ca","last_reissued_at":"2026-05-18T00:16:03.959503Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:16:03.959503Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1805.05062","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":"2026-05-18T00:16:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kK6mylNWcUbVvWFurkhggFqoeM5Boxnn3M279YkJokRV5yTa+GKkTRo2flc6Gym/PM+h+89OzENMmOMThpayBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T00:17:24.728297Z"},"content_sha256":"2b629e9c0e702a96063254db6cc27f833e79b1869d6677893005f2ce68471a34","schema_version":"1.0","event_id":"sha256:2b629e9c0e702a96063254db6cc27f833e79b1869d6677893005f2ce68471a34"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:QPVLLC37DCCC5MTP56I2FUUI5V","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Token-level and sequence-level loss smoothing for RNN language models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.CL","authors_text":"Jakob Verbeek, Laurent Besacier, Maha Elbayad","submitted_at":"2018-05-14T08:37:50Z","abstract_excerpt":"Despite the effectiveness of recurrent neural network language models, their maximum likelihood estimation suffers from two limitations. It treats all sentences that do not match the ground truth as equally poor, ignoring the structure of the output space. Second, it suffers from \"exposure bias\": during training tokens are predicted given ground-truth sequences, while at test time prediction is conditioned on generated output sequences. To overcome these limitations we build upon the recent reward augmented maximum likelihood approach \\ie sequence-level smoothing that encourages the model to p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1805.05062","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":"2026-05-18T00:16:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fcjnet8Mm+O32kWem005NRBs56byoj8tXFkNF3ds8ybH7F71od4U6n3uemibWLjZEakylPbu5pd9a7aWJJ2wBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-02T00:17:24.728654Z"},"content_sha256":"0737b7103d9479bc491265eb610f9ad9dd9302a874ba8b4f0faaf9900e037c43","schema_version":"1.0","event_id":"sha256:0737b7103d9479bc491265eb610f9ad9dd9302a874ba8b4f0faaf9900e037c43"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QPVLLC37DCCC5MTP56I2FUUI5V/bundle.json","state_url":"https://pith.science/pith/QPVLLC37DCCC5MTP56I2FUUI5V/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QPVLLC37DCCC5MTP56I2FUUI5V/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":"2026-06-02T00:17:24Z","links":{"resolver":"https://pith.science/pith/QPVLLC37DCCC5MTP56I2FUUI5V","bundle":"https://pith.science/pith/QPVLLC37DCCC5MTP56I2FUUI5V/bundle.json","state":"https://pith.science/pith/QPVLLC37DCCC5MTP56I2FUUI5V/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QPVLLC37DCCC5MTP56I2FUUI5V/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:QPVLLC37DCCC5MTP56I2FUUI5V","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":"d7c5b89134b17ddaeba2ea71e0870cae74f6cac89be524aff865241b1db2cf1c","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-05-14T08:37:50Z","title_canon_sha256":"a0fc414edb0769912df155183fb4f68dd2d6c8fb1b6ea331ebca94437832420f"},"schema_version":"1.0","source":{"id":"1805.05062","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1805.05062","created_at":"2026-05-18T00:16:03Z"},{"alias_kind":"arxiv_version","alias_value":"1805.05062v1","created_at":"2026-05-18T00:16:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1805.05062","created_at":"2026-05-18T00:16:03Z"},{"alias_kind":"pith_short_12","alias_value":"QPVLLC37DCCC","created_at":"2026-05-18T12:32:46Z"},{"alias_kind":"pith_short_16","alias_value":"QPVLLC37DCCC5MTP","created_at":"2026-05-18T12:32:46Z"},{"alias_kind":"pith_short_8","alias_value":"QPVLLC37","created_at":"2026-05-18T12:32:46Z"}],"graph_snapshots":[{"event_id":"sha256:0737b7103d9479bc491265eb610f9ad9dd9302a874ba8b4f0faaf9900e037c43","target":"graph","created_at":"2026-05-18T00:16:03Z","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":"Despite the effectiveness of recurrent neural network language models, their maximum likelihood estimation suffers from two limitations. It treats all sentences that do not match the ground truth as equally poor, ignoring the structure of the output space. Second, it suffers from \"exposure bias\": during training tokens are predicted given ground-truth sequences, while at test time prediction is conditioned on generated output sequences. To overcome these limitations we build upon the recent reward augmented maximum likelihood approach \\ie sequence-level smoothing that encourages the model to p","authors_text":"Jakob Verbeek, Laurent Besacier, Maha Elbayad","cross_cats":["cs.CV"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-05-14T08:37:50Z","title":"Token-level and sequence-level loss smoothing for RNN language models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1805.05062","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":"sha256:2b629e9c0e702a96063254db6cc27f833e79b1869d6677893005f2ce68471a34","target":"record","created_at":"2026-05-18T00:16:03Z","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":"d7c5b89134b17ddaeba2ea71e0870cae74f6cac89be524aff865241b1db2cf1c","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-05-14T08:37:50Z","title_canon_sha256":"a0fc414edb0769912df155183fb4f68dd2d6c8fb1b6ea331ebca94437832420f"},"schema_version":"1.0","source":{"id":"1805.05062","kind":"arxiv","version":1}},"canonical_sha256":"83eab58b7f18842eb26fef91a2d288ed4e2e2d1ad115740fd9326779b42658ca","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"83eab58b7f18842eb26fef91a2d288ed4e2e2d1ad115740fd9326779b42658ca","first_computed_at":"2026-05-18T00:16:03.959503Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:16:03.959503Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Gzan9tQKj0vrJ2YvyPbaUyQQ9CI1D9lsd8vhfKLUsSu9RKYqzrhPxSKjknc5gkp+7Fys1jC/D+42gBS3WwSLAA==","signature_status":"signed_v1","signed_at":"2026-05-18T00:16:03.960018Z","signed_message":"canonical_sha256_bytes"},"source_id":"1805.05062","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2b629e9c0e702a96063254db6cc27f833e79b1869d6677893005f2ce68471a34","sha256:0737b7103d9479bc491265eb610f9ad9dd9302a874ba8b4f0faaf9900e037c43"],"state_sha256":"0123745d101addc8280a0c8287f8d5b52ff6312530276fa39385b6df17f7f03c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"52Y+8eJC/1ds7Jlwz1Pw40nRP9X1BYkRYt1cPUkyQhdwF4jOTZEDp00TuUGwuSyxnuHBcQ/TeEsZC+pufeanBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-02T00:17:24.730600Z","bundle_sha256":"7660919b5c116afd4334232fa817978dbd132765ff7487ef23adc699cd98779e"}}