{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:T5FEL7FDKYXPUNL5X53YOYOD37","short_pith_number":"pith:T5FEL7FD","schema_version":"1.0","canonical_sha256":"9f4a45fca3562efa357dbf778761c3dfc5018df148743133a47dad5017fddf8e","source":{"kind":"arxiv","id":"2406.17272","version":1},"attestation_state":"computed","paper":{"title":"A Comprehensive Solution to Connect Speech Encoder and Large Language Model for ASR","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jun Zhang, Lu Lu, Tao Han, Van Tung Pham, Wei Li, Yist Lin, Yuxuan Wang","submitted_at":"2024-06-25T04:35:50Z","abstract_excerpt":"Recent works have shown promising results in connecting speech encoders to large language models (LLMs) for speech recognition. However, several limitations persist, including limited fine-tuning options, a lack of mechanisms to enforce speech-text alignment, and high insertion errors especially in domain mismatch conditions. This paper presents a comprehensive solution to address these issues. We begin by investigating more thoughtful fine-tuning schemes. Next, we propose a matching loss to enhance alignment between modalities. Finally, we explore training and inference methods to mitigate hi"},"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":"2406.17272","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-25T04:35:50Z","cross_cats_sorted":[],"title_canon_sha256":"3a63867b2fa9230fbf70b178baf20f940491a8e428969a36186523b327277c14","abstract_canon_sha256":"992f81b1e7b35267b0e6b9d5ec87b9ef977559f124ec1af9287fe0283158cb6b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:36:32.051379Z","signature_b64":"KOdHJ3yK7A++64nPOFTugmlhGNhdKPLYb+dOiULljKnRZWkfGCW9CitjsAKyAivHiY9CuwH6NaCqmvolrXz6BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9f4a45fca3562efa357dbf778761c3dfc5018df148743133a47dad5017fddf8e","last_reissued_at":"2026-07-05T08:36:32.050941Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:36:32.050941Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Comprehensive Solution to Connect Speech Encoder and Large Language Model for ASR","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jun Zhang, Lu Lu, Tao Han, Van Tung Pham, Wei Li, Yist Lin, Yuxuan Wang","submitted_at":"2024-06-25T04:35:50Z","abstract_excerpt":"Recent works have shown promising results in connecting speech encoders to large language models (LLMs) for speech recognition. However, several limitations persist, including limited fine-tuning options, a lack of mechanisms to enforce speech-text alignment, and high insertion errors especially in domain mismatch conditions. This paper presents a comprehensive solution to address these issues. We begin by investigating more thoughtful fine-tuning schemes. Next, we propose a matching loss to enhance alignment between modalities. Finally, we explore training and inference methods to mitigate hi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.17272","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/2406.17272/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":"2406.17272","created_at":"2026-07-05T08:36:32.050998+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.17272v1","created_at":"2026-07-05T08:36:32.050998+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.17272","created_at":"2026-07-05T08:36:32.050998+00:00"},{"alias_kind":"pith_short_12","alias_value":"T5FEL7FDKYXP","created_at":"2026-07-05T08:36:32.050998+00:00"},{"alias_kind":"pith_short_16","alias_value":"T5FEL7FDKYXPUNL5","created_at":"2026-07-05T08:36:32.050998+00:00"},{"alias_kind":"pith_short_8","alias_value":"T5FEL7FD","created_at":"2026-07-05T08:36:32.050998+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.09332","citing_title":"Phonemes vs. Projectors: An Investigation of Speech-Language Interfaces for LLM-based ASR","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37","json":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37.json","graph_json":"https://pith.science/api/pith-number/T5FEL7FDKYXPUNL5X53YOYOD37/graph.json","events_json":"https://pith.science/api/pith-number/T5FEL7FDKYXPUNL5X53YOYOD37/events.json","paper":"https://pith.science/paper/T5FEL7FD"},"agent_actions":{"view_html":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37","download_json":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37.json","view_paper":"https://pith.science/paper/T5FEL7FD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.17272&json=true","fetch_graph":"https://pith.science/api/pith-number/T5FEL7FDKYXPUNL5X53YOYOD37/graph.json","fetch_events":"https://pith.science/api/pith-number/T5FEL7FDKYXPUNL5X53YOYOD37/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37/action/timestamp_anchor","attest_storage":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37/action/storage_attestation","attest_author":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37/action/author_attestation","sign_citation":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37/action/citation_signature","submit_replication":"https://pith.science/pith/T5FEL7FDKYXPUNL5X53YOYOD37/action/replication_record"}},"created_at":"2026-07-05T08:36:32.050998+00:00","updated_at":"2026-07-05T08:36:32.050998+00:00"}