{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:WYMWHKUK7G5ROYZU5G3VRM6VZ5","short_pith_number":"pith:WYMWHKUK","schema_version":"1.0","canonical_sha256":"b61963aa8af9bb176334e9b758b3d5cf7763b016a41d9fa6f0c5be2abb6eb721","source":{"kind":"arxiv","id":"2305.13286","version":2},"attestation_state":"computed","paper":{"title":"How do languages influence each other? Studying cross-lingual data sharing during LM fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Garrette, Ekaterina Shutova, Rochelle Choenni","submitted_at":"2023-05-22T17:47:41Z","abstract_excerpt":"Multilingual large language models (MLLMs) are jointly trained on data from many different languages such that representation of individual languages can benefit from other languages' data. Impressive performance on zero-shot cross-lingual transfer shows that these models are capable of exploiting data from other languages. Yet, it remains unclear to what extent, and under which conditions, languages rely on each other's data. In this study, we use TracIn (Pruthi et al., 2020), a training data attribution (TDA) method, to retrieve the most influential training samples seen during multilingual "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2305.13286","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-22T17:47:41Z","cross_cats_sorted":[],"title_canon_sha256":"ae0ae41497079ddac5c4d66dcf4834a90094a314f8ea1942649e5ec9d22bd122","abstract_canon_sha256":"840df29cc14e0a71e12764bfc63ce185c8b10004a8f7a90ad1cf44ff572b941f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:21:04.969812Z","signature_b64":"zqRnDL8ZCJz965u0KKvBQqwoRhlzCn4Z6gKjEavig6MQ9SAH6I9Ws4a55jCdXmwAJOa/ivfPSxx1uIcb0ze7Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b61963aa8af9bb176334e9b758b3d5cf7763b016a41d9fa6f0c5be2abb6eb721","last_reissued_at":"2026-07-05T08:21:04.969377Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:21:04.969377Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How do languages influence each other? Studying cross-lingual data sharing during LM fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Garrette, Ekaterina Shutova, Rochelle Choenni","submitted_at":"2023-05-22T17:47:41Z","abstract_excerpt":"Multilingual large language models (MLLMs) are jointly trained on data from many different languages such that representation of individual languages can benefit from other languages' data. Impressive performance on zero-shot cross-lingual transfer shows that these models are capable of exploiting data from other languages. Yet, it remains unclear to what extent, and under which conditions, languages rely on each other's data. In this study, we use TracIn (Pruthi et al., 2020), a training data attribution (TDA) method, to retrieve the most influential training samples seen during multilingual "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.13286","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2305.13286/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2305.13286","created_at":"2026-07-05T08:21:04.969434+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.13286v2","created_at":"2026-07-05T08:21:04.969434+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.13286","created_at":"2026-07-05T08:21:04.969434+00:00"},{"alias_kind":"pith_short_12","alias_value":"WYMWHKUK7G5R","created_at":"2026-07-05T08:21:04.969434+00:00"},{"alias_kind":"pith_short_16","alias_value":"WYMWHKUK7G5ROYZU","created_at":"2026-07-05T08:21:04.969434+00:00"},{"alias_kind":"pith_short_8","alias_value":"WYMWHKUK","created_at":"2026-07-05T08:21:04.969434+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5","json":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5.json","graph_json":"https://pith.science/api/pith-number/WYMWHKUK7G5ROYZU5G3VRM6VZ5/graph.json","events_json":"https://pith.science/api/pith-number/WYMWHKUK7G5ROYZU5G3VRM6VZ5/events.json","paper":"https://pith.science/paper/WYMWHKUK"},"agent_actions":{"view_html":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5","download_json":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5.json","view_paper":"https://pith.science/paper/WYMWHKUK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.13286&json=true","fetch_graph":"https://pith.science/api/pith-number/WYMWHKUK7G5ROYZU5G3VRM6VZ5/graph.json","fetch_events":"https://pith.science/api/pith-number/WYMWHKUK7G5ROYZU5G3VRM6VZ5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5/action/storage_attestation","attest_author":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5/action/author_attestation","sign_citation":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5/action/citation_signature","submit_replication":"https://pith.science/pith/WYMWHKUK7G5ROYZU5G3VRM6VZ5/action/replication_record"}},"created_at":"2026-07-05T08:21:04.969434+00:00","updated_at":"2026-07-05T08:21:04.969434+00:00"}