{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:OTW5BAIDM7XK4CABTAZXWZFNUN","short_pith_number":"pith:OTW5BAID","canonical_record":{"source":{"id":"2106.08226","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-06-15T15:35:44Z","cross_cats_sorted":[],"title_canon_sha256":"66ff18e15c8a9822e9be771afa6f0a875cf43d78e8290cdc74426ad0d4adf861","abstract_canon_sha256":"e57138b7146619d192a164855a7f96e622e8d536c25902c0b85376b002686b32"},"schema_version":"1.0"},"canonical_sha256":"74edd0810367eeae080198337b64ada345928795cc3613a84609a5e47a3ef3ce","source":{"kind":"arxiv","id":"2106.08226","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.08226","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"arxiv_version","alias_value":"2106.08226v1","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.08226","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"pith_short_12","alias_value":"OTW5BAIDM7XK","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"pith_short_16","alias_value":"OTW5BAIDM7XK4CAB","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"pith_short_8","alias_value":"OTW5BAID","created_at":"2026-07-05T02:49:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:OTW5BAIDM7XK4CABTAZXWZFNUN","target":"record","payload":{"canonical_record":{"source":{"id":"2106.08226","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-06-15T15:35:44Z","cross_cats_sorted":[],"title_canon_sha256":"66ff18e15c8a9822e9be771afa6f0a875cf43d78e8290cdc74426ad0d4adf861","abstract_canon_sha256":"e57138b7146619d192a164855a7f96e622e8d536c25902c0b85376b002686b32"},"schema_version":"1.0"},"canonical_sha256":"74edd0810367eeae080198337b64ada345928795cc3613a84609a5e47a3ef3ce","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:49:38.680003Z","signature_b64":"kMhzPWydynSd3ljveQag2G+yHMF/2IQqmwJ02bmJsqvBgG2TkuHQOn5XJ4qD64NR8uRlxhkM39++Mpn5fdRCCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"74edd0810367eeae080198337b64ada345928795cc3613a84609a5e47a3ef3ce","last_reissued_at":"2026-07-05T02:49:38.679603Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:49:38.679603Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2106.08226","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":"2026-07-05T02:49:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hLl5JcVncglhG6Ags7hgCAV4W5pvYdDqpgqmS+UgMvuk0fbf0h8V5qmRuZLiaXMnUVfdcDByI07Abr5uhcE/Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T14:53:30.520721Z"},"content_sha256":"5a532cc69752feb7a46fde5a276c21d15cad2c45de56fdb75d194beca125f681","schema_version":"1.0","event_id":"sha256:5a532cc69752feb7a46fde5a276c21d15cad2c45de56fdb75d194beca125f681"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:OTW5BAIDM7XK4CABTAZXWZFNUN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Consistency Regularization for Cross-Lingual Fine-Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bo Zheng, Furu Wei, Li Dong, Saksham Singhal, Shaohan Huang, Ting Liu, Wanxiang Che, Wenhui Wang, Xia Song, Zewen Chi","submitted_at":"2021-06-15T15:35:44Z","abstract_excerpt":"Fine-tuning pre-trained cross-lingual language models can transfer task-specific supervision from one language to the others. In this work, we propose to improve cross-lingual fine-tuning with consistency regularization. Specifically, we use example consistency regularization to penalize the prediction sensitivity to four types of data augmentations, i.e., subword sampling, Gaussian noise, code-switch substitution, and machine translation. In addition, we employ model consistency to regularize the models trained with two augmented versions of the same training set. Experimental results on the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.08226","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2106.08226/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":"2026-07-05T02:49:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"b8AMz9HJs/ZGu59obCXmKoubjSAmwsDz5E2ezlML7cN44nxX8qIuAhcxRFXqQjhUzaCPSS/EY0sZH9hSEYP3CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-10T14:53:30.521098Z"},"content_sha256":"370487be3dd1588cd806ad98df1b756e9cce3c08c3d2de51101afdc32f47e30b","schema_version":"1.0","event_id":"sha256:370487be3dd1588cd806ad98df1b756e9cce3c08c3d2de51101afdc32f47e30b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OTW5BAIDM7XK4CABTAZXWZFNUN/bundle.json","state_url":"https://pith.science/pith/OTW5BAIDM7XK4CABTAZXWZFNUN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OTW5BAIDM7XK4CABTAZXWZFNUN/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":"2026-07-10T14:53:30Z","links":{"resolver":"https://pith.science/pith/OTW5BAIDM7XK4CABTAZXWZFNUN","bundle":"https://pith.science/pith/OTW5BAIDM7XK4CABTAZXWZFNUN/bundle.json","state":"https://pith.science/pith/OTW5BAIDM7XK4CABTAZXWZFNUN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OTW5BAIDM7XK4CABTAZXWZFNUN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:OTW5BAIDM7XK4CABTAZXWZFNUN","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":"e57138b7146619d192a164855a7f96e622e8d536c25902c0b85376b002686b32","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-06-15T15:35:44Z","title_canon_sha256":"66ff18e15c8a9822e9be771afa6f0a875cf43d78e8290cdc74426ad0d4adf861"},"schema_version":"1.0","source":{"id":"2106.08226","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2106.08226","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"arxiv_version","alias_value":"2106.08226v1","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.08226","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"pith_short_12","alias_value":"OTW5BAIDM7XK","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"pith_short_16","alias_value":"OTW5BAIDM7XK4CAB","created_at":"2026-07-05T02:49:38Z"},{"alias_kind":"pith_short_8","alias_value":"OTW5BAID","created_at":"2026-07-05T02:49:38Z"}],"graph_snapshots":[{"event_id":"sha256:370487be3dd1588cd806ad98df1b756e9cce3c08c3d2de51101afdc32f47e30b","target":"graph","created_at":"2026-07-05T02:49:38Z","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/2106.08226/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Fine-tuning pre-trained cross-lingual language models can transfer task-specific supervision from one language to the others. In this work, we propose to improve cross-lingual fine-tuning with consistency regularization. Specifically, we use example consistency regularization to penalize the prediction sensitivity to four types of data augmentations, i.e., subword sampling, Gaussian noise, code-switch substitution, and machine translation. In addition, we employ model consistency to regularize the models trained with two augmented versions of the same training set. Experimental results on the ","authors_text":"Bo Zheng, Furu Wei, Li Dong, Saksham Singhal, Shaohan Huang, Ting Liu, Wanxiang Che, Wenhui Wang, Xia Song, Zewen Chi","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-06-15T15:35:44Z","title":"Consistency Regularization for Cross-Lingual Fine-Tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.08226","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":"sha256:5a532cc69752feb7a46fde5a276c21d15cad2c45de56fdb75d194beca125f681","target":"record","created_at":"2026-07-05T02:49:38Z","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":"e57138b7146619d192a164855a7f96e622e8d536c25902c0b85376b002686b32","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-06-15T15:35:44Z","title_canon_sha256":"66ff18e15c8a9822e9be771afa6f0a875cf43d78e8290cdc74426ad0d4adf861"},"schema_version":"1.0","source":{"id":"2106.08226","kind":"arxiv","version":1}},"canonical_sha256":"74edd0810367eeae080198337b64ada345928795cc3613a84609a5e47a3ef3ce","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"74edd0810367eeae080198337b64ada345928795cc3613a84609a5e47a3ef3ce","first_computed_at":"2026-07-05T02:49:38.679603Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:49:38.679603Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kMhzPWydynSd3ljveQag2G+yHMF/2IQqmwJ02bmJsqvBgG2TkuHQOn5XJ4qD64NR8uRlxhkM39++Mpn5fdRCCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T02:49:38.680003Z","signed_message":"canonical_sha256_bytes"},"source_id":"2106.08226","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5a532cc69752feb7a46fde5a276c21d15cad2c45de56fdb75d194beca125f681","sha256:370487be3dd1588cd806ad98df1b756e9cce3c08c3d2de51101afdc32f47e30b"],"state_sha256":"f130d182909d91345f5aed63a0e85bfa9defcf2d30e57eea0634b86e55f375e8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VZNiiM6PdbtAXtDnLKiZX7Q+7eKatHIjgCd9mtTNHvtFQDEXV6NEczqbCan4I7sdaPBJqvdkbNm7+VF05rYBDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-10T14:53:30.523016Z","bundle_sha256":"1fdce3410aa8ca9d353e7abbf6de6ffcd576e22339ccd639e21b1b7d4502c8be"}}