{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LMUD5STECEDSL3X62ATQRN3W4A","short_pith_number":"pith:LMUD5STE","schema_version":"1.0","canonical_sha256":"5b283eca64110725eefed02708b776e037ea0d6abe4b8dae7adf9c1aceb26663","source":{"kind":"arxiv","id":"2504.08685","version":2},"attestation_state":"computed","paper":{"title":"Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Ceyuan Yang, Fangyuan Kong, Feilong Zuo, Fei Xiao, Feng Cheng, Feng Ling, Hao Chen, Haoyuan Guo, Heng Zhang, Houmin Wei, Huafeng Kuang, Huixia Li, Jerry Duncan, Jianchao Yang, Jiangqiao Yan, Jiashi Feng, Jiashi Li, Junda Zhang, Junru Zheng, Liangke Gui, Li Sun, Lu Jiang, Lu Qi, Manlin Zhang, Meng Wei, Peihao Zhu, Qi Zhao, Renfei Sun, Rui Wang, Sen Wang, Shanchuan Lin, Sheng Bi, Shu Liu, Siyu Zhang, Team Seawead, Tuyen Hoang, Xiaobin Zhuang, Xiaojie Li, Xin Xia, Xuefeng Xiao, Xuejiao Zeng, Xuyan Chi, Yanghua Peng, Yang Zhao, Yuping Wang, Yuxi Ren, Yuxuan Wang, Zhenheng Yang, Zhibei Ma, Zhijie Lin, Zhiwu Qing, Zhongkai Zhao, Zhuo Chen, Ziyan Yang, Zuquan Song","submitted_at":"2025-04-11T16:46:20Z","abstract_excerpt":"This technical report presents a cost-efficient strategy for training a video generation foundation model. We present a mid-sized research model with approximately 7 billion parameters (7B) called Seaweed-7B trained from scratch using 665,000 H100 GPU hours. Despite being trained with moderate computational resources, Seaweed-7B demonstrates highly competitive performance compared to contemporary video generation models of much larger size. Design choices are especially crucial in a resource-constrained setting. This technical report highlights the key design decisions that enhance the perform"},"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":"2504.08685","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-11T16:46:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1bba01bcd067a929e51c53f3707ac78b090308abb8455330cce9f8d51509560c","abstract_canon_sha256":"edb11fe3a84757c688f6c27097fba23b0d6e9b369bbf50bfff35bc3a71b14094"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:58:25.911864Z","signature_b64":"qC3aNVhZYftUT4OdMlHNrKPMNd5VfPaQo7nn1RDeIOC9U+QjT7/8/5uUcVZd4n4ArEVfQHijM+Oi6pI/EU/JDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b283eca64110725eefed02708b776e037ea0d6abe4b8dae7adf9c1aceb26663","last_reissued_at":"2026-07-05T10:58:25.911370Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:58:25.911370Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Ceyuan Yang, Fangyuan Kong, Feilong Zuo, Fei Xiao, Feng Cheng, Feng Ling, Hao Chen, Haoyuan Guo, Heng Zhang, Houmin Wei, Huafeng Kuang, Huixia Li, Jerry Duncan, Jianchao Yang, Jiangqiao Yan, Jiashi Feng, Jiashi Li, Junda Zhang, Junru Zheng, Liangke Gui, Li Sun, Lu Jiang, Lu Qi, Manlin Zhang, Meng Wei, Peihao Zhu, Qi Zhao, Renfei Sun, Rui Wang, Sen Wang, Shanchuan Lin, Sheng Bi, Shu Liu, Siyu Zhang, Team Seawead, Tuyen Hoang, Xiaobin Zhuang, Xiaojie Li, Xin Xia, Xuefeng Xiao, Xuejiao Zeng, Xuyan Chi, Yanghua Peng, Yang Zhao, Yuping Wang, Yuxi Ren, Yuxuan Wang, Zhenheng Yang, Zhibei Ma, Zhijie Lin, Zhiwu Qing, Zhongkai Zhao, Zhuo Chen, Ziyan Yang, Zuquan Song","submitted_at":"2025-04-11T16:46:20Z","abstract_excerpt":"This technical report presents a cost-efficient strategy for training a video generation foundation model. We present a mid-sized research model with approximately 7 billion parameters (7B) called Seaweed-7B trained from scratch using 665,000 H100 GPU hours. Despite being trained with moderate computational resources, Seaweed-7B demonstrates highly competitive performance compared to contemporary video generation models of much larger size. Design choices are especially crucial in a resource-constrained setting. This technical report highlights the key design decisions that enhance the perform"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.08685","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/2504.08685/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":"2504.08685","created_at":"2026-07-05T10:58:25.911429+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.08685v2","created_at":"2026-07-05T10:58:25.911429+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.08685","created_at":"2026-07-05T10:58:25.911429+00:00"},{"alias_kind":"pith_short_12","alias_value":"LMUD5STECEDS","created_at":"2026-07-05T10:58:25.911429+00:00"},{"alias_kind":"pith_short_16","alias_value":"LMUD5STECEDSL3X6","created_at":"2026-07-05T10:58:25.911429+00:00"},{"alias_kind":"pith_short_8","alias_value":"LMUD5STE","created_at":"2026-07-05T10:58:25.911429+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":15,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.13289","citing_title":"HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers","ref_index":193,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27736","citing_title":"Explicit Critic Guidance for Aligning Diffusion Models","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23610","citing_title":"EM-Vid: Training-Free Entity-Centric Memory for Efficient and Consistent Multi-Shot Video Generation","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2601.04068","citing_title":"Mind the Generative Details: Direct Localized Detail Preference Optimization for Video Diffusion Models","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27505","citing_title":"Leveraging Verifier-Based Reinforcement Learning in Image Editing","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2601.04068","citing_title":"Mind the Generative Details: Direct Localized Detail Preference Optimization for Video Diffusion Models","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2509.25161","citing_title":"Rolling Forcing: Autoregressive Long Video Diffusion in Real Time","ref_index":89,"is_internal_anchor":false},{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":93,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27505","citing_title":"Leveraging Verifier-Based Reinforcement Learning in Image Editing","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2506.09113","citing_title":"Seedance 1.0: Exploring the Boundaries of Video Generation Models","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11521","citing_title":"Continuous Adversarial Flow Models","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06939","citing_title":"Grounded Forcing: Bridging Time-Independent Semantics and Proximal Dynamics in Autoregressive Video Synthesis","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06339","citing_title":"Evolution of Video Generative Foundations","ref_index":160,"is_internal_anchor":false},{"citing_arxiv_id":"2505.14683","citing_title":"Emerging Properties in Unified Multimodal Pretraining","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14148","citing_title":"Seedance 2.0: Advancing Video Generation for World Complexity","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A","json":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A.json","graph_json":"https://pith.science/api/pith-number/LMUD5STECEDSL3X62ATQRN3W4A/graph.json","events_json":"https://pith.science/api/pith-number/LMUD5STECEDSL3X62ATQRN3W4A/events.json","paper":"https://pith.science/paper/LMUD5STE"},"agent_actions":{"view_html":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A","download_json":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A.json","view_paper":"https://pith.science/paper/LMUD5STE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.08685&json=true","fetch_graph":"https://pith.science/api/pith-number/LMUD5STECEDSL3X62ATQRN3W4A/graph.json","fetch_events":"https://pith.science/api/pith-number/LMUD5STECEDSL3X62ATQRN3W4A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A/action/storage_attestation","attest_author":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A/action/author_attestation","sign_citation":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A/action/citation_signature","submit_replication":"https://pith.science/pith/LMUD5STECEDSL3X62ATQRN3W4A/action/replication_record"}},"created_at":"2026-07-05T10:58:25.911429+00:00","updated_at":"2026-07-05T10:58:25.911429+00:00"}