{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:QQU53S3B77SLRSDT4X37MQ2AL7","short_pith_number":"pith:QQU53S3B","canonical_record":{"source":{"id":"2204.08110","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-04-17T23:56:54Z","cross_cats_sorted":[],"title_canon_sha256":"71486b0f5ff32dfac3878d540189eb877ef6c2c241124fc390fb62dbc137210b","abstract_canon_sha256":"e1507ef5e5236e001f1af13624ebfccb984607370e9d6151ada43c65e6a7a4c2"},"schema_version":"1.0"},"canonical_sha256":"8429ddcb61ffe4b8c873e5f7f643405fec4f64f128663eacbb4e13a38b8b067f","source":{"kind":"arxiv","id":"2204.08110","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2204.08110","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"arxiv_version","alias_value":"2204.08110v4","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2204.08110","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"pith_short_12","alias_value":"QQU53S3B77SL","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"pith_short_16","alias_value":"QQU53S3B77SLRSDT","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"pith_short_8","alias_value":"QQU53S3B","created_at":"2026-07-05T05:16:50Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:QQU53S3B77SLRSDT4X37MQ2AL7","target":"record","payload":{"canonical_record":{"source":{"id":"2204.08110","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-04-17T23:56:54Z","cross_cats_sorted":[],"title_canon_sha256":"71486b0f5ff32dfac3878d540189eb877ef6c2c241124fc390fb62dbc137210b","abstract_canon_sha256":"e1507ef5e5236e001f1af13624ebfccb984607370e9d6151ada43c65e6a7a4c2"},"schema_version":"1.0"},"canonical_sha256":"8429ddcb61ffe4b8c873e5f7f643405fec4f64f128663eacbb4e13a38b8b067f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:16:50.631463Z","signature_b64":"LljL0f54lNbD8lsu/VvMuti31U/WuO8S1Yni+NOawpZsZ/VQNLzuiPFdM/0yQf/KlfMK8zv8hm1Y9KSmZ62FAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8429ddcb61ffe4b8c873e5f7f643405fec4f64f128663eacbb4e13a38b8b067f","last_reissued_at":"2026-07-05T05:16:50.630977Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:16:50.630977Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2204.08110","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:16:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ufZ9GdmFWsaFDGsHCgmb05+qNoGUiv+zoovZNzfmLb8aCCubFM93IalUM0yOWuq9oXU27WsFWY3kpQcYQxGsAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T10:30:08.552337Z"},"content_sha256":"269acb972357714cc46f86850c0882dfbb952f42f61e616d89b85ab3dbec7765","schema_version":"1.0","event_id":"sha256:269acb972357714cc46f86850c0882dfbb952f42f61e616d89b85ab3dbec7765"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:QQU53S3B77SLRSDT4X37MQ2AL7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Language Contamination Helps Explain the Cross-lingual Capabilities of English Pretrained Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Luke Zettlemoyer, Terra Blevins","submitted_at":"2022-04-17T23:56:54Z","abstract_excerpt":"English pretrained language models, which make up the backbone of many modern NLP systems, require huge amounts of unlabeled training data. These models are generally presented as being trained only on English text but have been found to transfer surprisingly well to other languages. We investigate this phenomenon and find that common English pretraining corpora actually contain significant amounts of non-English text: even when less than 1% of data is not English (well within the error rate of strong language classifiers), this leads to hundreds of millions of foreign language tokens in large"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2204.08110","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2204.08110/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"2026-07-05T05:16:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"W3TQg/Yx7kqh+hwdZ0N/6EWBoFITxIxKLlSM2YvYB2bZskOfXWTg8Bbz7ot+WtdDwHA0o77Zrt+7CcrUC5J/AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T10:30:08.552714Z"},"content_sha256":"3966f9580d8d0dc3ee4ce7090e820315147ef31bff15e520d2c46363ff028a9e","schema_version":"1.0","event_id":"sha256:3966f9580d8d0dc3ee4ce7090e820315147ef31bff15e520d2c46363ff028a9e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QQU53S3B77SLRSDT4X37MQ2AL7/bundle.json","state_url":"https://pith.science/pith/QQU53S3B77SLRSDT4X37MQ2AL7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QQU53S3B77SLRSDT4X37MQ2AL7/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":"2026-07-07T10:30:08Z","links":{"resolver":"https://pith.science/pith/QQU53S3B77SLRSDT4X37MQ2AL7","bundle":"https://pith.science/pith/QQU53S3B77SLRSDT4X37MQ2AL7/bundle.json","state":"https://pith.science/pith/QQU53S3B77SLRSDT4X37MQ2AL7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QQU53S3B77SLRSDT4X37MQ2AL7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:QQU53S3B77SLRSDT4X37MQ2AL7","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":"e1507ef5e5236e001f1af13624ebfccb984607370e9d6151ada43c65e6a7a4c2","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-04-17T23:56:54Z","title_canon_sha256":"71486b0f5ff32dfac3878d540189eb877ef6c2c241124fc390fb62dbc137210b"},"schema_version":"1.0","source":{"id":"2204.08110","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2204.08110","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"arxiv_version","alias_value":"2204.08110v4","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2204.08110","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"pith_short_12","alias_value":"QQU53S3B77SL","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"pith_short_16","alias_value":"QQU53S3B77SLRSDT","created_at":"2026-07-05T05:16:50Z"},{"alias_kind":"pith_short_8","alias_value":"QQU53S3B","created_at":"2026-07-05T05:16:50Z"}],"graph_snapshots":[{"event_id":"sha256:3966f9580d8d0dc3ee4ce7090e820315147ef31bff15e520d2c46363ff028a9e","target":"graph","created_at":"2026-07-05T05:16:50Z","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"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2204.08110/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"English pretrained language models, which make up the backbone of many modern NLP systems, require huge amounts of unlabeled training data. These models are generally presented as being trained only on English text but have been found to transfer surprisingly well to other languages. We investigate this phenomenon and find that common English pretraining corpora actually contain significant amounts of non-English text: even when less than 1% of data is not English (well within the error rate of strong language classifiers), this leads to hundreds of millions of foreign language tokens in large","authors_text":"Luke Zettlemoyer, Terra Blevins","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-04-17T23:56:54Z","title":"Language Contamination Helps Explain the Cross-lingual Capabilities of English Pretrained Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2204.08110","kind":"arxiv","version":4},"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":"sha256:269acb972357714cc46f86850c0882dfbb952f42f61e616d89b85ab3dbec7765","target":"record","created_at":"2026-07-05T05:16:50Z","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":"e1507ef5e5236e001f1af13624ebfccb984607370e9d6151ada43c65e6a7a4c2","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-04-17T23:56:54Z","title_canon_sha256":"71486b0f5ff32dfac3878d540189eb877ef6c2c241124fc390fb62dbc137210b"},"schema_version":"1.0","source":{"id":"2204.08110","kind":"arxiv","version":4}},"canonical_sha256":"8429ddcb61ffe4b8c873e5f7f643405fec4f64f128663eacbb4e13a38b8b067f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8429ddcb61ffe4b8c873e5f7f643405fec4f64f128663eacbb4e13a38b8b067f","first_computed_at":"2026-07-05T05:16:50.630977Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:16:50.630977Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"LljL0f54lNbD8lsu/VvMuti31U/WuO8S1Yni+NOawpZsZ/VQNLzuiPFdM/0yQf/KlfMK8zv8hm1Y9KSmZ62FAw==","signature_status":"signed_v1","signed_at":"2026-07-05T05:16:50.631463Z","signed_message":"canonical_sha256_bytes"},"source_id":"2204.08110","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:269acb972357714cc46f86850c0882dfbb952f42f61e616d89b85ab3dbec7765","sha256:3966f9580d8d0dc3ee4ce7090e820315147ef31bff15e520d2c46363ff028a9e"],"state_sha256":"35770035f3da68c6d332889f912b8da31f53081a4b1e5485d561cdc6ba2bf7ff"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EDz2S0vLhUG2++lqJW2ry8etXzps9206Rcp1/0rEvqItgjTt6Gv2peH6ZmPhyKhuGlKw8ptg2s5CB8OmC/oNBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T10:30:08.554719Z","bundle_sha256":"4b2ab8816a541b23cd74c6f9013be41de9ec0c91acd2e7be63b242e470be5a59"}}