{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:26FL364KLM7T2WNDSCCH5QUZ6A","short_pith_number":"pith:26FL364K","schema_version":"1.0","canonical_sha256":"d78abdfb8a5b3f3d59a390847ec299f00fc227b1053f1fcd304854180bc52145","source":{"kind":"arxiv","id":"2503.13319","version":2},"attestation_state":"computed","paper":{"title":"MagicDistillation: Weak-to-Strong Video Distillation for Large-Scale Few-Step Synthesis","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Daquan Zhou, Hanzhong Guo, Hongwei Yi, Michael Lingelbach, Shitong Shao, Tian Ye, Zeke Xie, Zhiqiang Xu","submitted_at":"2025-03-17T15:58:27Z","abstract_excerpt":"Recently, open-source video diffusion models (VDMs), such as WanX, Magic141 and HunyuanVideo, have been scaled to over 10 billion parameters. These large-scale VDMs have demonstrated significant improvements over smaller-scale VDMs across multiple dimensions, including enhanced visual quality and more natural motion dynamics. However, these models face two major limitations: (1) High inference overhead: Large-scale VDMs require approximately 10 minutes to synthesize a 28-step video on a single H100 GPU. (2) Limited in portrait video synthesis: Models like WanX-I2V and HunyuanVideo-I2V often pr"},"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":"2503.13319","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-17T15:58:27Z","cross_cats_sorted":[],"title_canon_sha256":"8b261bc7596998cd0d6fdb6e958e95dbea81e8c36b87c848a5a79ac9dc6a14cd","abstract_canon_sha256":"2cf3b20bc3cea2761dc8f71dd906e784ab9ba41c7a04941f29c453adc81af5c1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:53.628518Z","signature_b64":"i07Rqb/CrLFMM2cE3QXDcOAhrf5ctouVutvMUCqlH5VY4C8IoIe5tZHPGtSEIqQMy7l3NN+A/GjsyYMpzjpoBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d78abdfb8a5b3f3d59a390847ec299f00fc227b1053f1fcd304854180bc52145","last_reissued_at":"2026-07-05T10:41:53.627845Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:53.627845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MagicDistillation: Weak-to-Strong Video Distillation for Large-Scale Few-Step Synthesis","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Daquan Zhou, Hanzhong Guo, Hongwei Yi, Michael Lingelbach, Shitong Shao, Tian Ye, Zeke Xie, Zhiqiang Xu","submitted_at":"2025-03-17T15:58:27Z","abstract_excerpt":"Recently, open-source video diffusion models (VDMs), such as WanX, Magic141 and HunyuanVideo, have been scaled to over 10 billion parameters. These large-scale VDMs have demonstrated significant improvements over smaller-scale VDMs across multiple dimensions, including enhanced visual quality and more natural motion dynamics. However, these models face two major limitations: (1) High inference overhead: Large-scale VDMs require approximately 10 minutes to synthesize a 28-step video on a single H100 GPU. (2) Limited in portrait video synthesis: Models like WanX-I2V and HunyuanVideo-I2V often pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.13319","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/2503.13319/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":"2503.13319","created_at":"2026-07-05T10:41:53.627969+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.13319v2","created_at":"2026-07-05T10:41:53.627969+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.13319","created_at":"2026-07-05T10:41:53.627969+00:00"},{"alias_kind":"pith_short_12","alias_value":"26FL364KLM7T","created_at":"2026-07-05T10:41:53.627969+00:00"},{"alias_kind":"pith_short_16","alias_value":"26FL364KLM7T2WND","created_at":"2026-07-05T10:41:53.627969+00:00"},{"alias_kind":"pith_short_8","alias_value":"26FL364K","created_at":"2026-07-05T10:41:53.627969+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17800","citing_title":"MaineCoon: Pursuing A Real-Time Audio-Visual Social World Model","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03972","citing_title":"AAD-1: Asymmetric Adversarial Distillation for One-Step Autoregressive Video Generation","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04569","citing_title":"LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25801","citing_title":"PixelWizard: Towards Efficient High-Fidelity Video Generation at Ultra-Large Spatial Resolution","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11804","citing_title":"OmniShow: Unifying Multimodal Conditions for Human-Object Interaction Video Generation","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15911","citing_title":"Efficient Video Diffusion Models: Advancements and Challenges","ref_index":114,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A","json":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A.json","graph_json":"https://pith.science/api/pith-number/26FL364KLM7T2WNDSCCH5QUZ6A/graph.json","events_json":"https://pith.science/api/pith-number/26FL364KLM7T2WNDSCCH5QUZ6A/events.json","paper":"https://pith.science/paper/26FL364K"},"agent_actions":{"view_html":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A","download_json":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A.json","view_paper":"https://pith.science/paper/26FL364K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.13319&json=true","fetch_graph":"https://pith.science/api/pith-number/26FL364KLM7T2WNDSCCH5QUZ6A/graph.json","fetch_events":"https://pith.science/api/pith-number/26FL364KLM7T2WNDSCCH5QUZ6A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A/action/storage_attestation","attest_author":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A/action/author_attestation","sign_citation":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A/action/citation_signature","submit_replication":"https://pith.science/pith/26FL364KLM7T2WNDSCCH5QUZ6A/action/replication_record"}},"created_at":"2026-07-05T10:41:53.627969+00:00","updated_at":"2026-07-05T10:41:53.627969+00:00"}