{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LJND2JKPBURDYMUMHZ3PR6PRRW","short_pith_number":"pith:LJND2JKP","schema_version":"1.0","canonical_sha256":"5a5a3d254f0d223c328c3e76f8f9f18da1820cb7857577a2284907f8a9fcfe1a","source":{"kind":"arxiv","id":"2606.21670","version":1},"attestation_state":"computed","paper":{"title":"Improving Text-to-Music Generation with Human Preference Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.SD","authors_text":"Chris Donahue, Haiwen Xia, Junwon Lee, Yinghao Ma, Yonghyun Kim","submitted_at":"2026-06-19T18:22:24Z","abstract_excerpt":"We describe our entry to the efficiency track of the Academic Text-to-Music (ATTM) Grand Challenge at ICME 2026. Beyond the challenge protocol's FAD-CLAP and CLAP score, we add a learned human-preference reward from TuneJury, a twin pairwise ranker trained over open music-preference datasets. The reward serves both as a training-time conditioning signal and as a sample-selection criterion. The pipeline combines five engineering decisions on a 120M-parameter FluxAudio-S backbone, four at training time and one at inference: (i) training-time reward conditioning that doubles as an inference-time "},"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":"2606.21670","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2026-06-19T18:22:24Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"315de81e5f258ce2363e7de41bf5dd712329ccd02838eacd4e23fecd261e61a0","abstract_canon_sha256":"5edaee6028260e47ae845354acdc06b3b19a106b0c56fa2657b85c4570604f9b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T01:13:18.501678Z","signature_b64":"WvDB+whEoyoiLGwijiIx3tlBcjWQJvzUnMx1FfhotZe41rYocSa45mRUJIyKflTPEgpkcKrbGBvmFJJ1K8hSBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a5a3d254f0d223c328c3e76f8f9f18da1820cb7857577a2284907f8a9fcfe1a","last_reissued_at":"2026-06-23T01:13:18.501116Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T01:13:18.501116Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Text-to-Music Generation with Human Preference Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.SD","authors_text":"Chris Donahue, Haiwen Xia, Junwon Lee, Yinghao Ma, Yonghyun Kim","submitted_at":"2026-06-19T18:22:24Z","abstract_excerpt":"We describe our entry to the efficiency track of the Academic Text-to-Music (ATTM) Grand Challenge at ICME 2026. Beyond the challenge protocol's FAD-CLAP and CLAP score, we add a learned human-preference reward from TuneJury, a twin pairwise ranker trained over open music-preference datasets. The reward serves both as a training-time conditioning signal and as a sample-selection criterion. The pipeline combines five engineering decisions on a 120M-parameter FluxAudio-S backbone, four at training time and one at inference: (i) training-time reward conditioning that doubles as an inference-time "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.21670","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/2606.21670/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":"2606.21670","created_at":"2026-06-23T01:13:18.501194+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.21670v1","created_at":"2026-06-23T01:13:18.501194+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.21670","created_at":"2026-06-23T01:13:18.501194+00:00"},{"alias_kind":"pith_short_12","alias_value":"LJND2JKPBURD","created_at":"2026-06-23T01:13:18.501194+00:00"},{"alias_kind":"pith_short_16","alias_value":"LJND2JKPBURDYMUM","created_at":"2026-06-23T01:13:18.501194+00:00"},{"alias_kind":"pith_short_8","alias_value":"LJND2JKP","created_at":"2026-06-23T01:13:18.501194+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/LJND2JKPBURDYMUMHZ3PR6PRRW","json":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW.json","graph_json":"https://pith.science/api/pith-number/LJND2JKPBURDYMUMHZ3PR6PRRW/graph.json","events_json":"https://pith.science/api/pith-number/LJND2JKPBURDYMUMHZ3PR6PRRW/events.json","paper":"https://pith.science/paper/LJND2JKP"},"agent_actions":{"view_html":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW","download_json":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW.json","view_paper":"https://pith.science/paper/LJND2JKP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.21670&json=true","fetch_graph":"https://pith.science/api/pith-number/LJND2JKPBURDYMUMHZ3PR6PRRW/graph.json","fetch_events":"https://pith.science/api/pith-number/LJND2JKPBURDYMUMHZ3PR6PRRW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW/action/storage_attestation","attest_author":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW/action/author_attestation","sign_citation":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW/action/citation_signature","submit_replication":"https://pith.science/pith/LJND2JKPBURDYMUMHZ3PR6PRRW/action/replication_record"}},"created_at":"2026-06-23T01:13:18.501194+00:00","updated_at":"2026-06-23T01:13:18.501194+00:00"}