{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:JANS4RF5KPVVSXU6GKJB3ZYFEW","short_pith_number":"pith:JANS4RF5","schema_version":"1.0","canonical_sha256":"481b2e44bd53eb595e9e32921de70525af6c06e855fc57ddc9e176115957fcdd","source":{"kind":"arxiv","id":"2305.15255","version":4},"attestation_state":"computed","paper":{"title":"Spoken Question Answering and Speech Continuation Using Spectrogram-Powered LLM","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Alon Levkovitch, Chulayuth Asawaroengchai, Ehud Rivlin, Eliya Nachmani, Julian Salazar, Michelle Tadmor Ramanovich, RJ Skerry-Ryan, Roy Hirsch, Soroosh Mariooryad","submitted_at":"2023-05-24T15:39:43Z","abstract_excerpt":"We present Spectron, a novel approach to adapting pre-trained large language models (LLMs) to perform spoken question answering (QA) and speech continuation. By endowing the LLM with a pre-trained speech encoder, our model becomes able to take speech inputs and generate speech outputs. The entire system is trained end-to-end and operates directly on spectrograms, simplifying our architecture. Key to our approach is a training objective that jointly supervises speech recognition, text continuation, and speech synthesis using only paired speech-text pairs, enabling a `cross-modal' chain-of-thoug"},"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":"2305.15255","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T15:39:43Z","cross_cats_sorted":["cs.LG","cs.SD","eess.AS"],"title_canon_sha256":"5727a237f694959494dad2378a73d67d961d592f2f79d891a97a1965140bb39c","abstract_canon_sha256":"427a443fe2f35617d716cb7852a187fcd28ba0ddfbe4c8649232d48589a136f5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:25:27.621395Z","signature_b64":"UgxVy6qS97KdhYIyy+5LbBK0yAVhvW2sRNR6dO0GMtYvjv6477vdU50mIsOrkIe9jXWcJncV8ox0YGuYsqkyCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"481b2e44bd53eb595e9e32921de70525af6c06e855fc57ddc9e176115957fcdd","last_reissued_at":"2026-07-05T08:25:27.620910Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:25:27.620910Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Spoken Question Answering and Speech Continuation Using Spectrogram-Powered LLM","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Alon Levkovitch, Chulayuth Asawaroengchai, Ehud Rivlin, Eliya Nachmani, Julian Salazar, Michelle Tadmor Ramanovich, RJ Skerry-Ryan, Roy Hirsch, Soroosh Mariooryad","submitted_at":"2023-05-24T15:39:43Z","abstract_excerpt":"We present Spectron, a novel approach to adapting pre-trained large language models (LLMs) to perform spoken question answering (QA) and speech continuation. By endowing the LLM with a pre-trained speech encoder, our model becomes able to take speech inputs and generate speech outputs. The entire system is trained end-to-end and operates directly on spectrograms, simplifying our architecture. Key to our approach is a training objective that jointly supervises speech recognition, text continuation, and speech synthesis using only paired speech-text pairs, enabling a `cross-modal' chain-of-thoug"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.15255","kind":"arxiv","version":4},"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/2305.15255/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":"2305.15255","created_at":"2026-07-05T08:25:27.620970+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.15255v4","created_at":"2026-07-05T08:25:27.620970+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.15255","created_at":"2026-07-05T08:25:27.620970+00:00"},{"alias_kind":"pith_short_12","alias_value":"JANS4RF5KPVV","created_at":"2026-07-05T08:25:27.620970+00:00"},{"alias_kind":"pith_short_16","alias_value":"JANS4RF5KPVVSXU6","created_at":"2026-07-05T08:25:27.620970+00:00"},{"alias_kind":"pith_short_8","alias_value":"JANS4RF5","created_at":"2026-07-05T08:25:27.620970+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12199","citing_title":"Which Speech Representation Better Matches Text-Native Reasoning? A Study of Speech-Text Alignment on Frame Rate and Representation","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05121","citing_title":"Audio Interaction Model","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30145","citing_title":"FacePlex: Full-Duplex Joint Speech-Facial Motion Generation for Conversational Avatars","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27190","citing_title":"Learning When to Think While Listening in Large Audio-Language Models","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2509.03526","citing_title":"Enhancing Speech Large Language Models through Reinforced Behavior Alignment","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2603.17837","citing_title":"The Silent Thought: Modeling Internal Cognition in Full-Duplex Spoken Dialogue Models via Latent Reasoning","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2603.17837","citing_title":"The Silent Thought: Modeling Internal Cognition in Full-Duplex Spoken Dialogue Models via Latent Reasoning","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06765","citing_title":"VITA-QinYu: Expressive Spoken Language Model for Role-Playing and Singing","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW","json":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW.json","graph_json":"https://pith.science/api/pith-number/JANS4RF5KPVVSXU6GKJB3ZYFEW/graph.json","events_json":"https://pith.science/api/pith-number/JANS4RF5KPVVSXU6GKJB3ZYFEW/events.json","paper":"https://pith.science/paper/JANS4RF5"},"agent_actions":{"view_html":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW","download_json":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW.json","view_paper":"https://pith.science/paper/JANS4RF5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.15255&json=true","fetch_graph":"https://pith.science/api/pith-number/JANS4RF5KPVVSXU6GKJB3ZYFEW/graph.json","fetch_events":"https://pith.science/api/pith-number/JANS4RF5KPVVSXU6GKJB3ZYFEW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW/action/storage_attestation","attest_author":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW/action/author_attestation","sign_citation":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW/action/citation_signature","submit_replication":"https://pith.science/pith/JANS4RF5KPVVSXU6GKJB3ZYFEW/action/replication_record"}},"created_at":"2026-07-05T08:25:27.620970+00:00","updated_at":"2026-07-05T08:25:27.620970+00:00"}