{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VO5PWDYVDBE5ZV74HYVOBLQECC","short_pith_number":"pith:VO5PWDYV","schema_version":"1.0","canonical_sha256":"abbafb0f151849dcd7fc3e2ae0ae0410a1112b8eb171609654752ab5c394b31b","source":{"kind":"arxiv","id":"2502.04896","version":2},"attestation_state":"computed","paper":{"title":"Goku: Flow Based Video Generative Foundation Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bingyue Peng, Chongjian Ge, Chuan Li, Fengda Zhu, Fu Li, Hao Yang, Hongxiang Hao, Hui Wu, Peize Sun, Ping Luo, Shilong Zhang, Shoufa Chen, Ting-Che Lin, Xiaobing Liu, Xing Wang, Yanghua Peng, Yida Zhang, Yifei Hu, Yi Jiang, Yuqi Zhang, Zehuan Yuan, Zhichao Lai","submitted_at":"2025-02-07T13:03:55Z","abstract_excerpt":"This paper introduces Goku, a state-of-the-art family of joint image-and-video generation models leveraging rectified flow Transformers to achieve industry-leading performance. We detail the foundational elements enabling high-quality visual generation, including the data curation pipeline, model architecture design, flow formulation, and advanced infrastructure for efficient and robust large-scale training. The Goku models demonstrate superior performance in both qualitative and quantitative evaluations, setting new benchmarks across major tasks. Specifically, Goku achieves 0.76 on GenEval an"},"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":"2502.04896","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-07T13:03:55Z","cross_cats_sorted":[],"title_canon_sha256":"ee945e2a17bbf3a34401bf811bdf2c5d08d605912351c0fc183209a65d3ad886","abstract_canon_sha256":"04a5ea6ee70ec5ca5d1bed6e44b5082dce47534e4a17cfa2f7cf98a0888e5184"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:12:04.222243Z","signature_b64":"oIUOabAuEmE0Z2UurtaVHPhY4ClM9CtSPBOCOYEOIeekNqAUa8bgQz+Q+iuKMZ3HQU+4ZEY+mWPVTAmeiOoMBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"abbafb0f151849dcd7fc3e2ae0ae0410a1112b8eb171609654752ab5c394b31b","last_reissued_at":"2026-07-05T10:12:04.221755Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:12:04.221755Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Goku: Flow Based Video Generative Foundation Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bingyue Peng, Chongjian Ge, Chuan Li, Fengda Zhu, Fu Li, Hao Yang, Hongxiang Hao, Hui Wu, Peize Sun, Ping Luo, Shilong Zhang, Shoufa Chen, Ting-Che Lin, Xiaobing Liu, Xing Wang, Yanghua Peng, Yida Zhang, Yifei Hu, Yi Jiang, Yuqi Zhang, Zehuan Yuan, Zhichao Lai","submitted_at":"2025-02-07T13:03:55Z","abstract_excerpt":"This paper introduces Goku, a state-of-the-art family of joint image-and-video generation models leveraging rectified flow Transformers to achieve industry-leading performance. We detail the foundational elements enabling high-quality visual generation, including the data curation pipeline, model architecture design, flow formulation, and advanced infrastructure for efficient and robust large-scale training. The Goku models demonstrate superior performance in both qualitative and quantitative evaluations, setting new benchmarks across major tasks. Specifically, Goku achieves 0.76 on GenEval an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.04896","kind":"arxiv","version":2},"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/2502.04896/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":"2502.04896","created_at":"2026-07-05T10:12:04.221814+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.04896v2","created_at":"2026-07-05T10:12:04.221814+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.04896","created_at":"2026-07-05T10:12:04.221814+00:00"},{"alias_kind":"pith_short_12","alias_value":"VO5PWDYVDBE5","created_at":"2026-07-05T10:12:04.221814+00:00"},{"alias_kind":"pith_short_16","alias_value":"VO5PWDYVDBE5ZV74","created_at":"2026-07-05T10:12:04.221814+00:00"},{"alias_kind":"pith_short_8","alias_value":"VO5PWDYV","created_at":"2026-07-05T10:12:04.221814+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.17991","citing_title":"Demystifying Transition Matching: When and Why It Can Beat Flow Matching","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2512.09646","citing_title":"VHOI: Controllable Video Generation of Human-Object Interactions from Sparse Trajectories via Motion Densification","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2503.07598","citing_title":"VACE: All-in-One Video Creation and Editing","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06339","citing_title":"Evolution of Video Generative Foundations","ref_index":84,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC","json":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC.json","graph_json":"https://pith.science/api/pith-number/VO5PWDYVDBE5ZV74HYVOBLQECC/graph.json","events_json":"https://pith.science/api/pith-number/VO5PWDYVDBE5ZV74HYVOBLQECC/events.json","paper":"https://pith.science/paper/VO5PWDYV"},"agent_actions":{"view_html":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC","download_json":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC.json","view_paper":"https://pith.science/paper/VO5PWDYV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.04896&json=true","fetch_graph":"https://pith.science/api/pith-number/VO5PWDYVDBE5ZV74HYVOBLQECC/graph.json","fetch_events":"https://pith.science/api/pith-number/VO5PWDYVDBE5ZV74HYVOBLQECC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC/action/storage_attestation","attest_author":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC/action/author_attestation","sign_citation":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC/action/citation_signature","submit_replication":"https://pith.science/pith/VO5PWDYVDBE5ZV74HYVOBLQECC/action/replication_record"}},"created_at":"2026-07-05T10:12:04.221814+00:00","updated_at":"2026-07-05T10:12:04.221814+00:00"}