{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:POEWCBW2UWGOTSETIWZIW4DY2V","short_pith_number":"pith:POEWCBW2","schema_version":"1.0","canonical_sha256":"7b896106daa58ce9c89345b28b7078d56f1a9ddfc074cfdbafad1d9683f1b10c","source":{"kind":"arxiv","id":"2505.10561","version":1},"attestation_state":"computed","paper":{"title":"T2A-Feedback: Improving Basic Capabilities of Text-to-Audio Generation via Fine-grained AI Feedback","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.SD","authors_text":"Chen Zhu, Jiawei Huang, Ke Lei, Luping Liu, Sashuai Zhou, Shengpeng Ji, Tao Jin, Xize Cheng, Zehan Wang, Zhenhui Ye, Zhou Zhao","submitted_at":"2025-05-15T17:59:29Z","abstract_excerpt":"Text-to-audio (T2A) generation has achieved remarkable progress in generating a variety of audio outputs from language prompts. However, current state-of-the-art T2A models still struggle to satisfy human preferences for prompt-following and acoustic quality when generating complex multi-event audio. To improve the performance of the model in these high-level applications, we propose to enhance the basic capabilities of the model with AI feedback learning. First, we introduce fine-grained AI audio scoring pipelines to: 1) verify whether each event in the text prompt is present in the audio (Ev"},"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":"2505.10561","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.SD","submitted_at":"2025-05-15T17:59:29Z","cross_cats_sorted":["eess.AS"],"title_canon_sha256":"9135475023c4c802e8b3875c90f1a1c257ef7f7dd69c1b5509eb814318bbca26","abstract_canon_sha256":"8288b35c3a560faa2d94b70437d3aad717b8c0c85affb5b9dbf6b40941220fb2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:03:45.734714Z","signature_b64":"Tg+DWXqNoGkJqBZqGlRYpDcwZ7lVruOt2IGu9QBLrJudisrtmg+vsloxUT1LWzaFkaGzPGvzsW5S5yGrjHqPAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7b896106daa58ce9c89345b28b7078d56f1a9ddfc074cfdbafad1d9683f1b10c","last_reissued_at":"2026-07-05T11:03:45.734092Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:03:45.734092Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"T2A-Feedback: Improving Basic Capabilities of Text-to-Audio Generation via Fine-grained AI Feedback","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.SD","authors_text":"Chen Zhu, Jiawei Huang, Ke Lei, Luping Liu, Sashuai Zhou, Shengpeng Ji, Tao Jin, Xize Cheng, Zehan Wang, Zhenhui Ye, Zhou Zhao","submitted_at":"2025-05-15T17:59:29Z","abstract_excerpt":"Text-to-audio (T2A) generation has achieved remarkable progress in generating a variety of audio outputs from language prompts. However, current state-of-the-art T2A models still struggle to satisfy human preferences for prompt-following and acoustic quality when generating complex multi-event audio. To improve the performance of the model in these high-level applications, we propose to enhance the basic capabilities of the model with AI feedback learning. First, we introduce fine-grained AI audio scoring pipelines to: 1) verify whether each event in the text prompt is present in the audio (Ev"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.10561","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/2505.10561/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":"2505.10561","created_at":"2026-07-05T11:03:45.734165+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.10561v1","created_at":"2026-07-05T11:03:45.734165+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.10561","created_at":"2026-07-05T11:03:45.734165+00:00"},{"alias_kind":"pith_short_12","alias_value":"POEWCBW2UWGO","created_at":"2026-07-05T11:03:45.734165+00:00"},{"alias_kind":"pith_short_16","alias_value":"POEWCBW2UWGOTSET","created_at":"2026-07-05T11:03:45.734165+00:00"},{"alias_kind":"pith_short_8","alias_value":"POEWCBW2","created_at":"2026-07-05T11:03:45.734165+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/POEWCBW2UWGOTSETIWZIW4DY2V","json":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V.json","graph_json":"https://pith.science/api/pith-number/POEWCBW2UWGOTSETIWZIW4DY2V/graph.json","events_json":"https://pith.science/api/pith-number/POEWCBW2UWGOTSETIWZIW4DY2V/events.json","paper":"https://pith.science/paper/POEWCBW2"},"agent_actions":{"view_html":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V","download_json":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V.json","view_paper":"https://pith.science/paper/POEWCBW2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.10561&json=true","fetch_graph":"https://pith.science/api/pith-number/POEWCBW2UWGOTSETIWZIW4DY2V/graph.json","fetch_events":"https://pith.science/api/pith-number/POEWCBW2UWGOTSETIWZIW4DY2V/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V/action/timestamp_anchor","attest_storage":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V/action/storage_attestation","attest_author":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V/action/author_attestation","sign_citation":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V/action/citation_signature","submit_replication":"https://pith.science/pith/POEWCBW2UWGOTSETIWZIW4DY2V/action/replication_record"}},"created_at":"2026-07-05T11:03:45.734165+00:00","updated_at":"2026-07-05T11:03:45.734165+00:00"}