{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:4IKLG75S3GZRL2BKTDVTHBV3UB","short_pith_number":"pith:4IKLG75S","schema_version":"1.0","canonical_sha256":"e214b37fb2d9b315e82a98eb3386bba07b165f8c3525b9810cb939a4df748b20","source":{"kind":"arxiv","id":"2309.13018","version":2},"attestation_state":"computed","paper":{"title":"Dynamic ASR Pathways: An Adaptive Masking Approach Towards Efficient Pruning of A Multilingual ASR Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Andros Tjandra, Chunyang Wu, Jay Mahadeokar, Jiamin Xie, Jinxi Guo, Junteng Jia, Ke Li, Leda Sari, Ozlem Kalinli, Yuan Shangguan","submitted_at":"2023-09-22T17:30:28Z","abstract_excerpt":"Neural network pruning offers an effective method for compressing a multilingual automatic speech recognition (ASR) model with minimal performance loss. However, it entails several rounds of pruning and re-training needed to be run for each language. In this work, we propose the use of an adaptive masking approach in two scenarios for pruning a multilingual ASR model efficiently, each resulting in sparse monolingual models or a sparse multilingual model (named as Dynamic ASR Pathways). Our approach dynamically adapts the sub-network, avoiding premature decisions about a fixed sub-network struc"},"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":"2309.13018","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2023-09-22T17:30:28Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG","cs.SD"],"title_canon_sha256":"c3cce1fbb4ed523007f517af446ae6db75e1b7e711ac0c2c838f14b22d417e51","abstract_canon_sha256":"42973e16cb27c9cd0f0e27123257352ff579195962de47e495fd2596b00112ce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:14.385086Z","signature_b64":"PRxaEYhy2BXYlyfRqHlDd4x8VSGMgpYStngoYGnqJ8z7KlDo7CpF0hd5MDeAZGiXGV9rBElJz46kk/X8EzLpDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e214b37fb2d9b315e82a98eb3386bba07b165f8c3525b9810cb939a4df748b20","last_reissued_at":"2026-07-05T11:23:14.384606Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:14.384606Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dynamic ASR Pathways: An Adaptive Masking Approach Towards Efficient Pruning of A Multilingual ASR Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Andros Tjandra, Chunyang Wu, Jay Mahadeokar, Jiamin Xie, Jinxi Guo, Junteng Jia, Ke Li, Leda Sari, Ozlem Kalinli, Yuan Shangguan","submitted_at":"2023-09-22T17:30:28Z","abstract_excerpt":"Neural network pruning offers an effective method for compressing a multilingual automatic speech recognition (ASR) model with minimal performance loss. However, it entails several rounds of pruning and re-training needed to be run for each language. In this work, we propose the use of an adaptive masking approach in two scenarios for pruning a multilingual ASR model efficiently, each resulting in sparse monolingual models or a sparse multilingual model (named as Dynamic ASR Pathways). Our approach dynamically adapts the sub-network, avoiding premature decisions about a fixed sub-network struc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.13018","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/2309.13018/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":"2309.13018","created_at":"2026-07-05T11:23:14.384668+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.13018v2","created_at":"2026-07-05T11:23:14.384668+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.13018","created_at":"2026-07-05T11:23:14.384668+00:00"},{"alias_kind":"pith_short_12","alias_value":"4IKLG75S3GZR","created_at":"2026-07-05T11:23:14.384668+00:00"},{"alias_kind":"pith_short_16","alias_value":"4IKLG75S3GZRL2BK","created_at":"2026-07-05T11:23:14.384668+00:00"},{"alias_kind":"pith_short_8","alias_value":"4IKLG75S","created_at":"2026-07-05T11:23:14.384668+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/4IKLG75S3GZRL2BKTDVTHBV3UB","json":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB.json","graph_json":"https://pith.science/api/pith-number/4IKLG75S3GZRL2BKTDVTHBV3UB/graph.json","events_json":"https://pith.science/api/pith-number/4IKLG75S3GZRL2BKTDVTHBV3UB/events.json","paper":"https://pith.science/paper/4IKLG75S"},"agent_actions":{"view_html":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB","download_json":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB.json","view_paper":"https://pith.science/paper/4IKLG75S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.13018&json=true","fetch_graph":"https://pith.science/api/pith-number/4IKLG75S3GZRL2BKTDVTHBV3UB/graph.json","fetch_events":"https://pith.science/api/pith-number/4IKLG75S3GZRL2BKTDVTHBV3UB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB/action/storage_attestation","attest_author":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB/action/author_attestation","sign_citation":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB/action/citation_signature","submit_replication":"https://pith.science/pith/4IKLG75S3GZRL2BKTDVTHBV3UB/action/replication_record"}},"created_at":"2026-07-05T11:23:14.384668+00:00","updated_at":"2026-07-05T11:23:14.384668+00:00"}