{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GCY27VB4DFOBVHEUDD74BB6DOI","short_pith_number":"pith:GCY27VB4","schema_version":"1.0","canonical_sha256":"30b1afd43c195c1a9c9418ffc087c3722a64cc7457d1cb3974769c80b9e848fa","source":{"kind":"arxiv","id":"2605.23261","version":1},"attestation_state":"computed","paper":{"title":"UniSRM: A Unified Speech Reward Model for Reasoning-Based Fine-grained Assessment","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Dongchao Yang, Helen Meng, Xixin Wu, Yayue Deng, Yiwen Guo, Yuanyuan Wang, Zhiyong Wu","submitted_at":"2026-05-22T06:02:09Z","abstract_excerpt":"Evaluating speech generation still relies heavily on human judgments, such as Mean Opinion Score (MOS), which are expensive, subjective, and difficult to reproduce at scale. While a few recent studies have begun to explore AudioLLM-based judge models, existing efforts typically target only a narrow set of scenarios (e.g., utterance-level quality or single-turn dialogue) and provide limited coverage of diverse speech generation tasks and evaluation dimensions. In this work, we propose UniSRM, a unified speech reward model that can support multi-dimensional, interpretable reward signals with rel"},"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":"2605.23261","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2026-05-22T06:02:09Z","cross_cats_sorted":["cs.SD"],"title_canon_sha256":"65bfae7533f19a41115dc65396ce425c146b8bbb2e2243cf304d6a6314478f05","abstract_canon_sha256":"697ef804c099421e87befd9afeb97412a5f7df8272c822d1186e37524d7131c0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:46.210708Z","signature_b64":"tL97BU3bBqHJhVj2EdtoqQhz8BN/s51xWfP7jGzc/6knvVK2EiIbjWPrn8MWZ3CI5/0NMJ6JL83E1if+vO8dAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"30b1afd43c195c1a9c9418ffc087c3722a64cc7457d1cb3974769c80b9e848fa","last_reissued_at":"2026-05-25T02:01:46.210059Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:46.210059Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"UniSRM: A Unified Speech Reward Model for Reasoning-Based Fine-grained Assessment","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Dongchao Yang, Helen Meng, Xixin Wu, Yayue Deng, Yiwen Guo, Yuanyuan Wang, Zhiyong Wu","submitted_at":"2026-05-22T06:02:09Z","abstract_excerpt":"Evaluating speech generation still relies heavily on human judgments, such as Mean Opinion Score (MOS), which are expensive, subjective, and difficult to reproduce at scale. While a few recent studies have begun to explore AudioLLM-based judge models, existing efforts typically target only a narrow set of scenarios (e.g., utterance-level quality or single-turn dialogue) and provide limited coverage of diverse speech generation tasks and evaluation dimensions. In this work, we propose UniSRM, a unified speech reward model that can support multi-dimensional, interpretable reward signals with rel"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.23261","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/2605.23261/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":"2605.23261","created_at":"2026-05-25T02:01:46.210179+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.23261v1","created_at":"2026-05-25T02:01:46.210179+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.23261","created_at":"2026-05-25T02:01:46.210179+00:00"},{"alias_kind":"pith_short_12","alias_value":"GCY27VB4DFOB","created_at":"2026-05-25T02:01:46.210179+00:00"},{"alias_kind":"pith_short_16","alias_value":"GCY27VB4DFOBVHEU","created_at":"2026-05-25T02:01:46.210179+00:00"},{"alias_kind":"pith_short_8","alias_value":"GCY27VB4","created_at":"2026-05-25T02:01:46.210179+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/GCY27VB4DFOBVHEUDD74BB6DOI","json":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI.json","graph_json":"https://pith.science/api/pith-number/GCY27VB4DFOBVHEUDD74BB6DOI/graph.json","events_json":"https://pith.science/api/pith-number/GCY27VB4DFOBVHEUDD74BB6DOI/events.json","paper":"https://pith.science/paper/GCY27VB4"},"agent_actions":{"view_html":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI","download_json":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI.json","view_paper":"https://pith.science/paper/GCY27VB4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.23261&json=true","fetch_graph":"https://pith.science/api/pith-number/GCY27VB4DFOBVHEUDD74BB6DOI/graph.json","fetch_events":"https://pith.science/api/pith-number/GCY27VB4DFOBVHEUDD74BB6DOI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI/action/storage_attestation","attest_author":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI/action/author_attestation","sign_citation":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI/action/citation_signature","submit_replication":"https://pith.science/pith/GCY27VB4DFOBVHEUDD74BB6DOI/action/replication_record"}},"created_at":"2026-05-25T02:01:46.210179+00:00","updated_at":"2026-05-25T02:01:46.210179+00:00"}