{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GTQKPWLZFJMODBOE6A2SFKAS5A","short_pith_number":"pith:GTQKPWLZ","schema_version":"1.0","canonical_sha256":"34e0a7d9792a58e185c4f03522a812e8398ed3ed1bd572d2e38a55ae8c0ffb25","source":{"kind":"arxiv","id":"2607.14962","version":1},"attestation_state":"computed","paper":{"title":"Multi-Axis Max@K Reinforcement Learning for Representative Diversity in Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Hiroki Furuta, Ku Onoda, Paavo Parmas, Shohei Taniguchi, Soichiro Nishimori, Yutaka Matsuo, Yuta Oshima","submitted_at":"2026-07-16T13:14:21Z","abstract_excerpt":"Text-to-image (T2I) models can synthesize realistic, prompt-aligned images, yet samples generated for the same prompt often cover only a small subset of visually distinct modes. This limits the diversity of images, and for person-centric prompts, can reflect or amplify demographic skew. We formalize this problem as coverage of a predefined set of semantically specified modes, which we call target-mode coverage. We then propose multi-axis max@K, a group-based reinforcement learning objective for improving such coverage in diffusion-based T2I models. Given a group of samples and one score per ta"},"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":"2607.14962","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-16T13:14:21Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"6bc6912aaeabec64e048fecf4f42814edf67a78414b1afd165699dded630fdc1","abstract_canon_sha256":"a69a69981dda5a2b18414758d66e66e34dbf7ff6deb01ca47f9455bbd4e75aad"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-17T01:22:01.260156Z","signature_b64":"FPrSljsG6BxnZxaEsRmftQ6QnTFj68LTqya8oHDlXR1I2xh7jeS4WcMH9HOYJbOd8TPCv/4nycElv4etOaqCCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34e0a7d9792a58e185c4f03522a812e8398ed3ed1bd572d2e38a55ae8c0ffb25","last_reissued_at":"2026-07-17T01:22:01.259312Z","signature_status":"signed_v1","first_computed_at":"2026-07-17T01:22:01.259312Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multi-Axis Max@K Reinforcement Learning for Representative Diversity in Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Hiroki Furuta, Ku Onoda, Paavo Parmas, Shohei Taniguchi, Soichiro Nishimori, Yutaka Matsuo, Yuta Oshima","submitted_at":"2026-07-16T13:14:21Z","abstract_excerpt":"Text-to-image (T2I) models can synthesize realistic, prompt-aligned images, yet samples generated for the same prompt often cover only a small subset of visually distinct modes. This limits the diversity of images, and for person-centric prompts, can reflect or amplify demographic skew. We formalize this problem as coverage of a predefined set of semantically specified modes, which we call target-mode coverage. We then propose multi-axis max@K, a group-based reinforcement learning objective for improving such coverage in diffusion-based T2I models. Given a group of samples and one score per ta"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.14962","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/2607.14962/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":"2607.14962","created_at":"2026-07-17T01:22:01.259752+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.14962v1","created_at":"2026-07-17T01:22:01.259752+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.14962","created_at":"2026-07-17T01:22:01.259752+00:00"},{"alias_kind":"pith_short_12","alias_value":"GTQKPWLZFJMO","created_at":"2026-07-17T01:22:01.259752+00:00"},{"alias_kind":"pith_short_16","alias_value":"GTQKPWLZFJMODBOE","created_at":"2026-07-17T01:22:01.259752+00:00"},{"alias_kind":"pith_short_8","alias_value":"GTQKPWLZ","created_at":"2026-07-17T01:22:01.259752+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/GTQKPWLZFJMODBOE6A2SFKAS5A","json":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A.json","graph_json":"https://pith.science/api/pith-number/GTQKPWLZFJMODBOE6A2SFKAS5A/graph.json","events_json":"https://pith.science/api/pith-number/GTQKPWLZFJMODBOE6A2SFKAS5A/events.json","paper":"https://pith.science/paper/GTQKPWLZ"},"agent_actions":{"view_html":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A","download_json":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A.json","view_paper":"https://pith.science/paper/GTQKPWLZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.14962&json=true","fetch_graph":"https://pith.science/api/pith-number/GTQKPWLZFJMODBOE6A2SFKAS5A/graph.json","fetch_events":"https://pith.science/api/pith-number/GTQKPWLZFJMODBOE6A2SFKAS5A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A/action/storage_attestation","attest_author":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A/action/author_attestation","sign_citation":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A/action/citation_signature","submit_replication":"https://pith.science/pith/GTQKPWLZFJMODBOE6A2SFKAS5A/action/replication_record"}},"created_at":"2026-07-17T01:22:01.259752+00:00","updated_at":"2026-07-17T01:22:01.259752+00:00"}