{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7EEHTPGT2IOV35SWBO4IVMYWDR","short_pith_number":"pith:7EEHTPGT","schema_version":"1.0","canonical_sha256":"f90879bcd3d21d5df6560bb88ab3161c638ef84dac21261500c0feb84de7e993","source":{"kind":"arxiv","id":"2606.28249","version":1},"attestation_state":"computed","paper":{"title":"HPRO: Hierarchical Progressive Reward Optimization via Preference Extraction for Emotional Text-to-Speech","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.SD"],"primary_cat":"eess.AS","authors_text":"Baiji Liu, Haoming Li, Jingyuan Xing, Ruitong Xiao, Rui Xing, Sihang Nie, Xiangmin Xu, Xiaofen Xing","submitted_at":"2026-06-26T16:35:48Z","abstract_excerpt":"Recently, Large Language Model (LLM)-based Text-to-Speech (TTS) models have achieved remarkable naturalness. However, the standard Supervised Fine-Tuning paradigm often converges to statistically averaged prosody, limiting emotional expressiveness. While preference-driven optimization offers a promising alternative, existing approaches suffer from two structural mismatches: information conflict, where content and emotion in a shared latent space produce conflicting gradients, leading to reward hacking and semantic degradation; and scale gap, where sparse sentence-level rewards struggle to guid"},"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":"2606.28249","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2026-06-26T16:35:48Z","cross_cats_sorted":["cs.CL","cs.SD"],"title_canon_sha256":"2fa3f8edf5ab01b4de4d99cd4a0a3315e029b72aba885f4f089a492e1ff38b6d","abstract_canon_sha256":"dbf9c83a159fed46d54766cce84bf89ac164281a962bbd8d26c1ffe6b80ae792"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T01:15:10.383993Z","signature_b64":"LjlH6hYlLv8iM3rfK5w+/PKA3cuGr1LGZxC6q99MzfC28QwtGtlhKi32bSr+hbf5xqY8OgLy3/MFCGs23SwcBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f90879bcd3d21d5df6560bb88ab3161c638ef84dac21261500c0feb84de7e993","last_reissued_at":"2026-06-29T01:15:10.383583Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T01:15:10.383583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HPRO: Hierarchical Progressive Reward Optimization via Preference Extraction for Emotional Text-to-Speech","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.SD"],"primary_cat":"eess.AS","authors_text":"Baiji Liu, Haoming Li, Jingyuan Xing, Ruitong Xiao, Rui Xing, Sihang Nie, Xiangmin Xu, Xiaofen Xing","submitted_at":"2026-06-26T16:35:48Z","abstract_excerpt":"Recently, Large Language Model (LLM)-based Text-to-Speech (TTS) models have achieved remarkable naturalness. However, the standard Supervised Fine-Tuning paradigm often converges to statistically averaged prosody, limiting emotional expressiveness. While preference-driven optimization offers a promising alternative, existing approaches suffer from two structural mismatches: information conflict, where content and emotion in a shared latent space produce conflicting gradients, leading to reward hacking and semantic degradation; and scale gap, where sparse sentence-level rewards struggle to guid"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.28249","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/2606.28249/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":"2606.28249","created_at":"2026-06-29T01:15:10.383646+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.28249v1","created_at":"2026-06-29T01:15:10.383646+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.28249","created_at":"2026-06-29T01:15:10.383646+00:00"},{"alias_kind":"pith_short_12","alias_value":"7EEHTPGT2IOV","created_at":"2026-06-29T01:15:10.383646+00:00"},{"alias_kind":"pith_short_16","alias_value":"7EEHTPGT2IOV35SW","created_at":"2026-06-29T01:15:10.383646+00:00"},{"alias_kind":"pith_short_8","alias_value":"7EEHTPGT","created_at":"2026-06-29T01:15:10.383646+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/7EEHTPGT2IOV35SWBO4IVMYWDR","json":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR.json","graph_json":"https://pith.science/api/pith-number/7EEHTPGT2IOV35SWBO4IVMYWDR/graph.json","events_json":"https://pith.science/api/pith-number/7EEHTPGT2IOV35SWBO4IVMYWDR/events.json","paper":"https://pith.science/paper/7EEHTPGT"},"agent_actions":{"view_html":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR","download_json":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR.json","view_paper":"https://pith.science/paper/7EEHTPGT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.28249&json=true","fetch_graph":"https://pith.science/api/pith-number/7EEHTPGT2IOV35SWBO4IVMYWDR/graph.json","fetch_events":"https://pith.science/api/pith-number/7EEHTPGT2IOV35SWBO4IVMYWDR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR/action/storage_attestation","attest_author":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR/action/author_attestation","sign_citation":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR/action/citation_signature","submit_replication":"https://pith.science/pith/7EEHTPGT2IOV35SWBO4IVMYWDR/action/replication_record"}},"created_at":"2026-06-29T01:15:10.383646+00:00","updated_at":"2026-06-29T01:15:10.383646+00:00"}