{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3SO5U6Y5OZHWP5E73HSNK4BMMD","short_pith_number":"pith:3SO5U6Y5","schema_version":"1.0","canonical_sha256":"dc9dda7b1d764f67f49fd9e4d5702c60d71a236811a97adaaf79dcff6a2d7b9f","source":{"kind":"arxiv","id":"2508.07014","version":2},"attestation_state":"computed","paper":{"title":"TurboBias: Universal ASR Context-Biasing powered by GPU-accelerated Phrase-Boosting Tree","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.SD"],"primary_cat":"eess.AS","authors_text":"Andrei Andrusenko, Boris Ginsburg, Lilit Grigoryan, Vitaly Lavrukhin, Vladimir Bataev","submitted_at":"2025-08-09T15:27:07Z","abstract_excerpt":"Recognizing specific key phrases is an essential task for contextualized Automatic Speech Recognition (ASR). However, most existing context-biasing approaches have limitations associated with the necessity of additional model training, significantly slow down the decoding process, or constrain the choice of the ASR system type. This paper proposes a universal ASR context-biasing framework that supports all major types: CTC, Transducers, and Attention Encoder-Decoder models. The framework is based on a GPU-accelerated word boosting tree, which enables it to be used in shallow fusion mode for gr"},"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":"2508.07014","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2025-08-09T15:27:07Z","cross_cats_sorted":["cs.AI","cs.CL","cs.SD"],"title_canon_sha256":"b9b78b5d1aedc83b50cdf716caad422db668883725334a197da5486a30b1e0d4","abstract_canon_sha256":"46c5f51c6bd53f5ed5d0d0c87f97810ff4b1cf9401ee1375137ea0ad5b64f449"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:52:24.159870Z","signature_b64":"r6xDY9fozEI8bdph+q4WLR7lKvS7dSJ9NX9+4v92YXc9Ho1ALEJajU236PEREa1IZEo1NXHJTmIYxEFA/OLoDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dc9dda7b1d764f67f49fd9e4d5702c60d71a236811a97adaaf79dcff6a2d7b9f","last_reissued_at":"2026-07-05T11:52:24.159392Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:52:24.159392Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TurboBias: Universal ASR Context-Biasing powered by GPU-accelerated Phrase-Boosting Tree","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.SD"],"primary_cat":"eess.AS","authors_text":"Andrei Andrusenko, Boris Ginsburg, Lilit Grigoryan, Vitaly Lavrukhin, Vladimir Bataev","submitted_at":"2025-08-09T15:27:07Z","abstract_excerpt":"Recognizing specific key phrases is an essential task for contextualized Automatic Speech Recognition (ASR). However, most existing context-biasing approaches have limitations associated with the necessity of additional model training, significantly slow down the decoding process, or constrain the choice of the ASR system type. This paper proposes a universal ASR context-biasing framework that supports all major types: CTC, Transducers, and Attention Encoder-Decoder models. The framework is based on a GPU-accelerated word boosting tree, which enables it to be used in shallow fusion mode for gr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.07014","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/2508.07014/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":"2508.07014","created_at":"2026-07-05T11:52:24.159454+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.07014v2","created_at":"2026-07-05T11:52:24.159454+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.07014","created_at":"2026-07-05T11:52:24.159454+00:00"},{"alias_kind":"pith_short_12","alias_value":"3SO5U6Y5OZHW","created_at":"2026-07-05T11:52:24.159454+00:00"},{"alias_kind":"pith_short_16","alias_value":"3SO5U6Y5OZHWP5E7","created_at":"2026-07-05T11:52:24.159454+00:00"},{"alias_kind":"pith_short_8","alias_value":"3SO5U6Y5","created_at":"2026-07-05T11:52:24.159454+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.07354","citing_title":"Contextual Earnings-22: A Speech Recognition Benchmark with Custom Vocabulary in the Wild","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD","json":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD.json","graph_json":"https://pith.science/api/pith-number/3SO5U6Y5OZHWP5E73HSNK4BMMD/graph.json","events_json":"https://pith.science/api/pith-number/3SO5U6Y5OZHWP5E73HSNK4BMMD/events.json","paper":"https://pith.science/paper/3SO5U6Y5"},"agent_actions":{"view_html":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD","download_json":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD.json","view_paper":"https://pith.science/paper/3SO5U6Y5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.07014&json=true","fetch_graph":"https://pith.science/api/pith-number/3SO5U6Y5OZHWP5E73HSNK4BMMD/graph.json","fetch_events":"https://pith.science/api/pith-number/3SO5U6Y5OZHWP5E73HSNK4BMMD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD/action/storage_attestation","attest_author":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD/action/author_attestation","sign_citation":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD/action/citation_signature","submit_replication":"https://pith.science/pith/3SO5U6Y5OZHWP5E73HSNK4BMMD/action/replication_record"}},"created_at":"2026-07-05T11:52:24.159454+00:00","updated_at":"2026-07-05T11:52:24.159454+00:00"}