{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SEGSO7LN26AUJLC5KOOSQI5N6L","short_pith_number":"pith:SEGSO7LN","schema_version":"1.0","canonical_sha256":"910d277d6dd78144ac5d539d2823adf2e136f6891f6ad168bd717ec124f90f51","source":{"kind":"arxiv","id":"2501.07875","version":1},"attestation_state":"computed","paper":{"title":"Continual Learning with Embedding Layer Surgery and Task-wise Beam Search using Whisper","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chin Yuen Kwok, Eng Siong Chng, Jia Qi Yip","submitted_at":"2025-01-14T06:33:40Z","abstract_excerpt":"Current Multilingual ASR models only support a fraction of the world's languages. Continual Learning (CL) aims to tackle this problem by adding new languages to pre-trained models while avoiding the loss of performance on existing languages, also known as Catastrophic Forgetting (CF). However, existing CL methods overlook the adaptation of the token embedding lookup table at the decoder, despite its significant contribution to CF. We propose Embedding Layer Surgery where separate copies of the token embeddings are created for each new languages, and one of the copies is selected to replace the"},"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":"2501.07875","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-14T06:33:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c634df047f290b060d160619fbedce881653c200d4640a3c57b76594ac540c27","abstract_canon_sha256":"de56b2e6171faf0a4c587f5006630f8d071a2b00e85c90d41fc7d2d81c7d7dce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:00:51.705244Z","signature_b64":"sfji6nmLGwKm/VGWPQlEzKQFH9ANNZGYs/fYuCyCYcNGGr0l8w4LYZySHkufGFu+RLDty2BZc8wYTh85RBs+Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"910d277d6dd78144ac5d539d2823adf2e136f6891f6ad168bd717ec124f90f51","last_reissued_at":"2026-07-05T10:00:51.704784Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:00:51.704784Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Continual Learning with Embedding Layer Surgery and Task-wise Beam Search using Whisper","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chin Yuen Kwok, Eng Siong Chng, Jia Qi Yip","submitted_at":"2025-01-14T06:33:40Z","abstract_excerpt":"Current Multilingual ASR models only support a fraction of the world's languages. Continual Learning (CL) aims to tackle this problem by adding new languages to pre-trained models while avoiding the loss of performance on existing languages, also known as Catastrophic Forgetting (CF). However, existing CL methods overlook the adaptation of the token embedding lookup table at the decoder, despite its significant contribution to CF. We propose Embedding Layer Surgery where separate copies of the token embeddings are created for each new languages, and one of the copies is selected to replace the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.07875","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/2501.07875/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":"2501.07875","created_at":"2026-07-05T10:00:51.704842+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.07875v1","created_at":"2026-07-05T10:00:51.704842+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.07875","created_at":"2026-07-05T10:00:51.704842+00:00"},{"alias_kind":"pith_short_12","alias_value":"SEGSO7LN26AU","created_at":"2026-07-05T10:00:51.704842+00:00"},{"alias_kind":"pith_short_16","alias_value":"SEGSO7LN26AUJLC5","created_at":"2026-07-05T10:00:51.704842+00:00"},{"alias_kind":"pith_short_8","alias_value":"SEGSO7LN","created_at":"2026-07-05T10:00:51.704842+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.24863","citing_title":"Rethinking Continual Learning for Speech and Audio: A Representation-Centric Taxonomy and Open Problems","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L","json":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L.json","graph_json":"https://pith.science/api/pith-number/SEGSO7LN26AUJLC5KOOSQI5N6L/graph.json","events_json":"https://pith.science/api/pith-number/SEGSO7LN26AUJLC5KOOSQI5N6L/events.json","paper":"https://pith.science/paper/SEGSO7LN"},"agent_actions":{"view_html":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L","download_json":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L.json","view_paper":"https://pith.science/paper/SEGSO7LN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.07875&json=true","fetch_graph":"https://pith.science/api/pith-number/SEGSO7LN26AUJLC5KOOSQI5N6L/graph.json","fetch_events":"https://pith.science/api/pith-number/SEGSO7LN26AUJLC5KOOSQI5N6L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L/action/storage_attestation","attest_author":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L/action/author_attestation","sign_citation":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L/action/citation_signature","submit_replication":"https://pith.science/pith/SEGSO7LN26AUJLC5KOOSQI5N6L/action/replication_record"}},"created_at":"2026-07-05T10:00:51.704842+00:00","updated_at":"2026-07-05T10:00:51.704842+00:00"}