{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "qwen-team-audio-taskconditioned-family-2023", "title": "Qwen-Audio: task-conditioned audio-language pretraining and natural-prompt interaction", "summary": "Developed hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation.", "boundary": "Inherits audio encoders, language models, multitask learning and DPO. One family package with explicit two-stage deltas, not two model-size entries. Qwen2-Audio outputs text responses; it is not the later Thinker/Talker speech-generation architecture. No universal audio understanding guarantee.", "area": "Audio-language systems", "topic": "Artificial intelligence", "year": 2023, "lane": "HUMAN", "result": "EMPIRICAL", "package_id": "qwen-team-audio-taskconditioned-family-2023", "target": { "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "content_hash": "sha256:0d47eb2452bfd22ba50efabbc771b9f5080c9dc88efdf4170cfa256a061123ec" }, "humans": [ "Yunfei Chu", "Jin Xu", "Xiaohuan Zhou", "Qian Yang", "Shiliang Zhang", "Zhijie Yan", "Chang Zhou", "Jingren Zhou", "Haojie Wei", "Xipin Wei", "Zhifang Guo", "Yichong Leng", "Yuanjun Lv", "Jinzheng He", "Junyang Lin" ], "authors_complete": true, "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.", "agents": [], "related": [], "subresults": [], "scientific_assessment": { "source_review": "SOURCE_CURATED", "scientific_review": "NOT_REVIEWED_BY_WRS", "formal_replay": "NOT_RUN", "provider_attestation": "NONE" }, "sources": [ { "id": "cur20261015-arxiv-2311-07919v1", "title": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" }, { "id": "cur20261015-arxiv-2407-10759v1", "title": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } ], "points": "110000", "published_grade": { "id": "qwen-team-audio-taskconditioned-family-2023", "entry_id": "qwen-team-audio-taskconditioned-family-2023", "band": "A", "modifier": null, "target": { "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "content_hash": "sha256:0d47eb2452bfd22ba50efabbc771b9f5080c9dc88efdf4170cfa256a061123ec" }, "what_changed": "Developed hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation.", "scope": "Developed hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation. Inherits audio encoders, language models, multitask learning and DPO. One family package with explicit two-stage deltas, not two model-size entries. Qwen2-Audio outputs text responses; it is not the later Thinker/Talker speech-generation architecture. No universal audio understanding guarantee.", "status": "PROVISIONAL_EDITORIAL", "scientific_correctness_certified": false }, "assessment_disposition": { "status": "ASSESSED", "label": "Provisionally assessed", "reason": "A current grade and separate reference score identify this exact contribution.", "target": { "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "content_hash": "sha256:0d47eb2452bfd22ba50efabbc771b9f5080c9dc88efdf4170cfa256a061123ec" }, "edition": "ratings-e039106f30dec177" } }, "version": { "kind": "PUBLIC_RECORD_VIEW", "previous": null, "published_at": "2026-10-16T00:00:00Z", "publisher": { "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33", "role": "WRS_EDITORIAL_PUBLICATION" }, "record": { "contributors": [ { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Yunfei Chu", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jin Xu", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Xiaohuan Zhou", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Qian Yang", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Shiliang Zhang", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zhijie Yan", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Chang Zhou", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jingren Zhou", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Haojie Wei", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Xipin Wei", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zhifang Guo", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Yichong Leng", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Yuanjun Lv", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jinzheng He", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Junyang Lin", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "discoverer" ], "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" } }, { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-1", "kind": "source_reference", "source": { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "Inherits audio encoders, language models, multitask learning and DPO. One family package with explicit two-stage deltas, not two model-size entries. Qwen2-Audio outputs text responses; it is not the later Thinker/Talker speech-generation architecture. No universal audio understanding guarantee.", "domain": "Audio-language systems" }, "sources": [ { "citation": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1" }, { "citation": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1" } ], "statement": "# Qwen-Audio: task-conditioned audio-language pretraining and natural-prompt interaction\n\nDeveloped hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation.\n\n## Scientific scope\n\nInherits audio encoders, language models, multitask learning and DPO. One family package with explicit two-stage deltas, not two model-size entries. Qwen2-Audio outputs text responses; it is not the later Thinker/Talker speech-generation architecture. No universal audio understanding guarantee.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nYunfei Chu; Jin Xu; Xiaohuan Zhou; Qian Yang; Shiliang Zhang; Zhijie Yan; Chang Zhou; Jingren Zhou; Haojie Wei; Xipin Wei; Zhifang Guo; Yichong Leng; Yuanjun Lv; Jinzheng He; Junyang Lin\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models](https://arxiv.org/abs/2311.07919v1)\n- [Qwen2-Audio Technical Report](https://arxiv.org/abs/2407.10759v1)\n", "summary": "Developed hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation.", "title": "Qwen-Audio: task-conditioned audio-language pretraining and natural-prompt interaction" }, "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "source_ref": { "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "content_hash": "sha256:0d47eb2452bfd22ba50efabbc771b9f5080c9dc88efdf4170cfa256a061123ec" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:aa1fe7dceb03e3e616d267a8d665d8e4c880281d9e94dde9b40abed4f128659b" }, "target": { "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "content_hash": "sha256:0d47eb2452bfd22ba50efabbc771b9f5080c9dc88efdf4170cfa256a061123ec" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "qwen-team-audio-taskconditioned-family-2023", "entry_id": "qwen-team-audio-taskconditioned-family-2023", "revision": 1, "title": "Qwen-Audio: task-conditioned audio-language pretraining and natural-prompt interaction", "context": "Audio-language systems", "track": "Human-designed research method or resource", "band": "A", "modifier": null, "status": "PROVISIONAL_EDITORIAL", "assessed_on": "2026-10-16", "evidence_cutoff": "2026-10-16", "achievement_state": "SUPPORTED_SCOPED_RESULT", "what_changed": "Developed hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation.", "scope": "Developed hierarchical task-tag conditioning for heterogeneous audio-language pretraining; the later Qwen2 stage replaced tags with natural-language task prompts and added audio-instruction preference adaptation. Inherits audio encoders, language models, multitask learning and DPO. One family package with explicit two-stage deltas, not two model-size entries. Qwen2-Audio outputs text responses; it is not the later Thinker/Talker speech-generation architecture. No universal audio understanding guarantee.", "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits audio encoders, language models, multitask learning and DPO. One family package with explicit two-stage deltas, not two model-size entries. Qwen2-Audio outputs text responses; it is not the later Thinker/Talker speech-generation architecture. No universal audio understanding guarantee.", "references": [ { "title": "Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Models", "url": "https://arxiv.org/abs/2311.07919v1", "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification." }, { "title": "Qwen2-Audio Technical Report", "url": "https://arxiv.org/abs/2407.10759v1", "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification." } ], "roles": [ { "who": "Source-listed authors and technical contributors", "credit": "SCIENCE", "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities." } ], "target": { "record_id": "urn:uuid:282b55a1-222e-464e-a621-ec08d952317b", "version": 1, "content_hash": "sha256:0d47eb2452bfd22ba50efabbc771b9f5080c9dc88efdf4170cfa256a061123ec" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "110000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": null, "reading_status": "SOURCE_LINKED_NOTE" }