{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5GCHJLDVN2WRYAOVWU5P2ZGYYU","short_pith_number":"pith:5GCHJLDV","schema_version":"1.0","canonical_sha256":"e98474ac756ead1c01d5b53afd64d8c510d85a64c71636188589678a0b129ebe","source":{"kind":"arxiv","id":"2410.03007","version":1},"attestation_state":"computed","paper":{"title":"FastAdaSP: Multitask-Adapted Efficient Inference for Large Speech Language Model","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"eess.AS","authors_text":"Chao-Han Huck Yang, Jiaqi Song, Shinji Watanabe, Yichen Lu","submitted_at":"2024-10-03T21:33:07Z","abstract_excerpt":"In this study, we aim to explore Multitask Speech Language Model (SpeechLM) efficient inference via token reduction. Unlike other modalities such as vision or text, speech has unique temporal dependencies, making previous efficient inference works on other modalities not directly applicable. Furthermore, methods for efficient SpeechLM inference on long sequence and sparse signals remain largely unexplored. Then we propose FastAdaSP, a weighted token merging framework specifically designed for various speech-related tasks to improve the trade-off between efficiency and performance. Experimental"},"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":"2410.03007","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"eess.AS","submitted_at":"2024-10-03T21:33:07Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"23530eb6bc3e235b302ac292232484eb7fb39bf57cd8171c0397af69462bef63","abstract_canon_sha256":"a3416971e8229b50bbbaa4c14ef86cda39988818e0f7d6ad17caaae043b18457"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:15:43.029652Z","signature_b64":"VvkvZsJmyF9Z/DTF8VP4MVPtSgE/bzHB+IPN/VWfKVDE+uTc8DIrkpqsqTqRFcN+s4/WgHa6ubJ26/iJul5cDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e98474ac756ead1c01d5b53afd64d8c510d85a64c71636188589678a0b129ebe","last_reissued_at":"2026-07-05T09:15:43.029230Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:15:43.029230Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FastAdaSP: Multitask-Adapted Efficient Inference for Large Speech Language Model","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"eess.AS","authors_text":"Chao-Han Huck Yang, Jiaqi Song, Shinji Watanabe, Yichen Lu","submitted_at":"2024-10-03T21:33:07Z","abstract_excerpt":"In this study, we aim to explore Multitask Speech Language Model (SpeechLM) efficient inference via token reduction. Unlike other modalities such as vision or text, speech has unique temporal dependencies, making previous efficient inference works on other modalities not directly applicable. Furthermore, methods for efficient SpeechLM inference on long sequence and sparse signals remain largely unexplored. Then we propose FastAdaSP, a weighted token merging framework specifically designed for various speech-related tasks to improve the trade-off between efficiency and performance. Experimental"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.03007","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/2410.03007/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":"2410.03007","created_at":"2026-07-05T09:15:43.029286+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.03007v1","created_at":"2026-07-05T09:15:43.029286+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.03007","created_at":"2026-07-05T09:15:43.029286+00:00"},{"alias_kind":"pith_short_12","alias_value":"5GCHJLDVN2WR","created_at":"2026-07-05T09:15:43.029286+00:00"},{"alias_kind":"pith_short_16","alias_value":"5GCHJLDVN2WRYAOV","created_at":"2026-07-05T09:15:43.029286+00:00"},{"alias_kind":"pith_short_8","alias_value":"5GCHJLDV","created_at":"2026-07-05T09:15:43.029286+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/5GCHJLDVN2WRYAOVWU5P2ZGYYU","json":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU.json","graph_json":"https://pith.science/api/pith-number/5GCHJLDVN2WRYAOVWU5P2ZGYYU/graph.json","events_json":"https://pith.science/api/pith-number/5GCHJLDVN2WRYAOVWU5P2ZGYYU/events.json","paper":"https://pith.science/paper/5GCHJLDV"},"agent_actions":{"view_html":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU","download_json":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU.json","view_paper":"https://pith.science/paper/5GCHJLDV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.03007&json=true","fetch_graph":"https://pith.science/api/pith-number/5GCHJLDVN2WRYAOVWU5P2ZGYYU/graph.json","fetch_events":"https://pith.science/api/pith-number/5GCHJLDVN2WRYAOVWU5P2ZGYYU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU/action/storage_attestation","attest_author":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU/action/author_attestation","sign_citation":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU/action/citation_signature","submit_replication":"https://pith.science/pith/5GCHJLDVN2WRYAOVWU5P2ZGYYU/action/replication_record"}},"created_at":"2026-07-05T09:15:43.029286+00:00","updated_at":"2026-07-05T09:15:43.029286+00:00"}