{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:MTUNM4IVMRTPF2F5CO5NAOSFTZ","short_pith_number":"pith:MTUNM4IV","canonical_record":{"source":{"id":"1808.09381","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-28T16:05:40Z","cross_cats_sorted":[],"title_canon_sha256":"8e5d035ade5a01172a90d4f2f7f05accfd0c81654407dd3443c31996a1b851a3","abstract_canon_sha256":"04564badb6c64bddc6570e1d2e4da7ca6a6ac4d5a651fbf3677e84d794c587f1"},"schema_version":"1.0"},"canonical_sha256":"64e8d671156466f2e8bd13bad03a459e74fb7b5739aa9409a3d80cfc67904908","source":{"kind":"arxiv","id":"1808.09381","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1808.09381","created_at":"2026-05-18T00:04:12Z"},{"alias_kind":"arxiv_version","alias_value":"1808.09381v2","created_at":"2026-05-18T00:04:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.09381","created_at":"2026-05-18T00:04:12Z"},{"alias_kind":"pith_short_12","alias_value":"MTUNM4IVMRTP","created_at":"2026-05-18T12:32:40Z"},{"alias_kind":"pith_short_16","alias_value":"MTUNM4IVMRTPF2F5","created_at":"2026-05-18T12:32:40Z"},{"alias_kind":"pith_short_8","alias_value":"MTUNM4IV","created_at":"2026-05-18T12:32:40Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:MTUNM4IVMRTPF2F5CO5NAOSFTZ","target":"record","payload":{"canonical_record":{"source":{"id":"1808.09381","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-28T16:05:40Z","cross_cats_sorted":[],"title_canon_sha256":"8e5d035ade5a01172a90d4f2f7f05accfd0c81654407dd3443c31996a1b851a3","abstract_canon_sha256":"04564badb6c64bddc6570e1d2e4da7ca6a6ac4d5a651fbf3677e84d794c587f1"},"schema_version":"1.0"},"canonical_sha256":"64e8d671156466f2e8bd13bad03a459e74fb7b5739aa9409a3d80cfc67904908","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:04:12.565381Z","signature_b64":"s06eMYtz88DTmiI9xBIvoZt9PkboQRyOJgYJJZxUVee23RxlgP220hwhUALpfLNQTzQyYhy4JkG+SAuqemGIDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64e8d671156466f2e8bd13bad03a459e74fb7b5739aa9409a3d80cfc67904908","last_reissued_at":"2026-05-18T00:04:12.564709Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:04:12.564709Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1808.09381","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":"2026-05-18T00:04:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vTDxCxihE/MD4Mfy7gxm2lkcs160SFCewaN4UZPsVSYVRw7INEY/+KgY7TiPTbHLCcMGHJ9jpExzU5F6S0XZBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T01:49:57.826309Z"},"content_sha256":"f511bc897ae6c641c057466ab0f8244503a1099b5a431b292cdd9f805179f44c","schema_version":"1.0","event_id":"sha256:f511bc897ae6c641c057466ab0f8244503a1099b5a431b292cdd9f805179f44c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:MTUNM4IVMRTPF2F5CO5NAOSFTZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Understanding Back-Translation at Scale","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"David Grangier, Michael Auli, Myle Ott, Sergey Edunov","submitted_at":"2018-08-28T16:05:40Z","abstract_excerpt":"An effective method to improve neural machine translation with monolingual data is to augment the parallel training corpus with back-translations of target language sentences. This work broadens the understanding of back-translation and investigates a number of methods to generate synthetic source sentences. We find that in all but resource poor settings back-translations obtained via sampling or noised beam outputs are most effective. Our analysis shows that sampling or noisy synthetic data gives a much stronger training signal than data generated by beam or greedy search. We also compare how"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.09381","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":"2026-05-18T00:04:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SwcUNEB/q8dyXkUUCfX47JXLbTtZK95g96tFTqLNj1Xq11Yi/h/obrheD1+W4Tak1JgoLBPgnDTfVQg90wbtAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T01:49:57.827031Z"},"content_sha256":"5a9a0c2b9efe3ea6e641d4e10aee0c56b81e69c7abc1032adfda71f6bc77e589","schema_version":"1.0","event_id":"sha256:5a9a0c2b9efe3ea6e641d4e10aee0c56b81e69c7abc1032adfda71f6bc77e589"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ/bundle.json","state_url":"https://pith.science/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ/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":"2026-06-09T01:49:57Z","links":{"resolver":"https://pith.science/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ","bundle":"https://pith.science/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ/bundle.json","state":"https://pith.science/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MTUNM4IVMRTPF2F5CO5NAOSFTZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:MTUNM4IVMRTPF2F5CO5NAOSFTZ","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":"04564badb6c64bddc6570e1d2e4da7ca6a6ac4d5a651fbf3677e84d794c587f1","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-28T16:05:40Z","title_canon_sha256":"8e5d035ade5a01172a90d4f2f7f05accfd0c81654407dd3443c31996a1b851a3"},"schema_version":"1.0","source":{"id":"1808.09381","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1808.09381","created_at":"2026-05-18T00:04:12Z"},{"alias_kind":"arxiv_version","alias_value":"1808.09381v2","created_at":"2026-05-18T00:04:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.09381","created_at":"2026-05-18T00:04:12Z"},{"alias_kind":"pith_short_12","alias_value":"MTUNM4IVMRTP","created_at":"2026-05-18T12:32:40Z"},{"alias_kind":"pith_short_16","alias_value":"MTUNM4IVMRTPF2F5","created_at":"2026-05-18T12:32:40Z"},{"alias_kind":"pith_short_8","alias_value":"MTUNM4IV","created_at":"2026-05-18T12:32:40Z"}],"graph_snapshots":[{"event_id":"sha256:5a9a0c2b9efe3ea6e641d4e10aee0c56b81e69c7abc1032adfda71f6bc77e589","target":"graph","created_at":"2026-05-18T00:04:12Z","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 effective method to improve neural machine translation with monolingual data is to augment the parallel training corpus with back-translations of target language sentences. This work broadens the understanding of back-translation and investigates a number of methods to generate synthetic source sentences. We find that in all but resource poor settings back-translations obtained via sampling or noised beam outputs are most effective. Our analysis shows that sampling or noisy synthetic data gives a much stronger training signal than data generated by beam or greedy search. We also compare how","authors_text":"David Grangier, Michael Auli, Myle Ott, Sergey Edunov","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-28T16:05:40Z","title":"Understanding Back-Translation at Scale"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.09381","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:f511bc897ae6c641c057466ab0f8244503a1099b5a431b292cdd9f805179f44c","target":"record","created_at":"2026-05-18T00:04:12Z","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":"04564badb6c64bddc6570e1d2e4da7ca6a6ac4d5a651fbf3677e84d794c587f1","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-28T16:05:40Z","title_canon_sha256":"8e5d035ade5a01172a90d4f2f7f05accfd0c81654407dd3443c31996a1b851a3"},"schema_version":"1.0","source":{"id":"1808.09381","kind":"arxiv","version":2}},"canonical_sha256":"64e8d671156466f2e8bd13bad03a459e74fb7b5739aa9409a3d80cfc67904908","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"64e8d671156466f2e8bd13bad03a459e74fb7b5739aa9409a3d80cfc67904908","first_computed_at":"2026-05-18T00:04:12.564709Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:04:12.564709Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"s06eMYtz88DTmiI9xBIvoZt9PkboQRyOJgYJJZxUVee23RxlgP220hwhUALpfLNQTzQyYhy4JkG+SAuqemGIDg==","signature_status":"signed_v1","signed_at":"2026-05-18T00:04:12.565381Z","signed_message":"canonical_sha256_bytes"},"source_id":"1808.09381","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f511bc897ae6c641c057466ab0f8244503a1099b5a431b292cdd9f805179f44c","sha256:5a9a0c2b9efe3ea6e641d4e10aee0c56b81e69c7abc1032adfda71f6bc77e589"],"state_sha256":"e3736df0c562cb7e07f2c13f99f4f1815baff9b7deecf77a688ea501cdc948d2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TjwqjCRJA4t9eXyI8B9YisWKmTKnyQo6o7IksNaAkHQE5epcwOqPF84uiOt4OMkvnpv1OI21IfqBSxoZPD+pBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-09T01:49:57.831051Z","bundle_sha256":"2e5fd03720be191b66d1f8454c799371cc5a91ccf68f9ea5498c3d13bc4585ea"}}