{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:GS4XC3W7S5BOXKZXCVBARSRUNO","short_pith_number":"pith:GS4XC3W7","schema_version":"1.0","canonical_sha256":"34b9716edf9742ebab37154208ca346b9bcbde8e78af6a77e172144ffbda5aba","source":{"kind":"arxiv","id":"2210.05709","version":1},"attestation_state":"computed","paper":{"title":"Shapley Head Pruning: Identifying and Removing Interference in Multilingual Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Diyi Yang, William Held","submitted_at":"2022-10-11T18:11:37Z","abstract_excerpt":"Multilingual transformer-based models demonstrate remarkable zero and few-shot transfer across languages by learning and reusing language-agnostic features. However, as a fixed-size model acquires more languages, its performance across all languages degrades, a phenomenon termed interference. Often attributed to limited model capacity, interference is commonly addressed by adding additional parameters despite evidence that transformer-based models are overparameterized. In this work, we show that it is possible to reduce interference by instead identifying and pruning language-specific paramet"},"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":"2210.05709","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-11T18:11:37Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"734f2589c531023a2cfbafa62e41ebdc18038701150af18c6cca2fc815a39527","abstract_canon_sha256":"cb39ee46037725218b319588c88404960582542cfdb5d547cfb04fb7a91613ad"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:07:09.128972Z","signature_b64":"+6AZCQPGRKztj+1tGXY8ntFXNudLjLc62RM19MQumV1mguF4Ew8J78KVmW3M4X3JB+EQuqKqMwgkWaQv2akFDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34b9716edf9742ebab37154208ca346b9bcbde8e78af6a77e172144ffbda5aba","last_reissued_at":"2026-07-05T06:07:09.128562Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:07:09.128562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Shapley Head Pruning: Identifying and Removing Interference in Multilingual Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Diyi Yang, William Held","submitted_at":"2022-10-11T18:11:37Z","abstract_excerpt":"Multilingual transformer-based models demonstrate remarkable zero and few-shot transfer across languages by learning and reusing language-agnostic features. However, as a fixed-size model acquires more languages, its performance across all languages degrades, a phenomenon termed interference. Often attributed to limited model capacity, interference is commonly addressed by adding additional parameters despite evidence that transformer-based models are overparameterized. In this work, we show that it is possible to reduce interference by instead identifying and pruning language-specific paramet"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.05709","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/2210.05709/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":"2210.05709","created_at":"2026-07-05T06:07:09.128627+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.05709v1","created_at":"2026-07-05T06:07:09.128627+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.05709","created_at":"2026-07-05T06:07:09.128627+00:00"},{"alias_kind":"pith_short_12","alias_value":"GS4XC3W7S5BO","created_at":"2026-07-05T06:07:09.128627+00:00"},{"alias_kind":"pith_short_16","alias_value":"GS4XC3W7S5BOXKZX","created_at":"2026-07-05T06:07:09.128627+00:00"},{"alias_kind":"pith_short_8","alias_value":"GS4XC3W7","created_at":"2026-07-05T06:07:09.128627+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.00743","citing_title":"Assortment of Attention Heads: Accelerating Federated PEFT with Head Pruning and Strategic Client Selection","ref_index":50,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO","json":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO.json","graph_json":"https://pith.science/api/pith-number/GS4XC3W7S5BOXKZXCVBARSRUNO/graph.json","events_json":"https://pith.science/api/pith-number/GS4XC3W7S5BOXKZXCVBARSRUNO/events.json","paper":"https://pith.science/paper/GS4XC3W7"},"agent_actions":{"view_html":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO","download_json":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO.json","view_paper":"https://pith.science/paper/GS4XC3W7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.05709&json=true","fetch_graph":"https://pith.science/api/pith-number/GS4XC3W7S5BOXKZXCVBARSRUNO/graph.json","fetch_events":"https://pith.science/api/pith-number/GS4XC3W7S5BOXKZXCVBARSRUNO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO/action/storage_attestation","attest_author":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO/action/author_attestation","sign_citation":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO/action/citation_signature","submit_replication":"https://pith.science/pith/GS4XC3W7S5BOXKZXCVBARSRUNO/action/replication_record"}},"created_at":"2026-07-05T06:07:09.128627+00:00","updated_at":"2026-07-05T06:07:09.128627+00:00"}