{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XHYEK46B7WBNPQA4HJBNSS2RT7","short_pith_number":"pith:XHYEK46B","schema_version":"1.0","canonical_sha256":"b9f04573c1fd82d7c01c3a42d94b519ffcabe0bc858d1ec57520632e3d389c0f","source":{"kind":"arxiv","id":"2503.14325","version":1},"attestation_state":"computed","paper":{"title":"LeanVAE: An Ultra-Efficient Reconstruction VAE for Video Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["eess.IV"],"primary_cat":"cs.CV","authors_text":"Fajie Yuan, Yu Cheng","submitted_at":"2025-03-18T14:58:59Z","abstract_excerpt":"Recent advances in Latent Video Diffusion Models (LVDMs) have revolutionized video generation by leveraging Video Variational Autoencoders (Video VAEs) to compress intricate video data into a compact latent space. However, as LVDM training scales, the computational overhead of Video VAEs becomes a critical bottleneck, particularly for encoding high-resolution videos. To address this, we propose LeanVAE, a novel and ultra-efficient Video VAE framework that introduces two key innovations: (1) a lightweight architecture based on a Neighborhood-Aware Feedforward (NAF) module and non-overlapping pa"},"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":"2503.14325","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-18T14:58:59Z","cross_cats_sorted":["eess.IV"],"title_canon_sha256":"1aa42787b5c860c4a0ef0f519d3f3b945c3d9515d49f65e4cb9cf18ed1ffee19","abstract_canon_sha256":"a51cdfd979f3a76cf8d4665c3ec8d19cde6cf27d67ca54bfeeaf60a54e7e7933"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:34:00.227894Z","signature_b64":"JUQJrnud5Ij3kjqjeZAxDh6Z2nWw7W4YsKtEetAG7BxD4FG/hUPNcUzgPXB34w5SfAdZeKuUOahPHEf/7TquDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b9f04573c1fd82d7c01c3a42d94b519ffcabe0bc858d1ec57520632e3d389c0f","last_reissued_at":"2026-07-05T10:34:00.227259Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:34:00.227259Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LeanVAE: An Ultra-Efficient Reconstruction VAE for Video Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["eess.IV"],"primary_cat":"cs.CV","authors_text":"Fajie Yuan, Yu Cheng","submitted_at":"2025-03-18T14:58:59Z","abstract_excerpt":"Recent advances in Latent Video Diffusion Models (LVDMs) have revolutionized video generation by leveraging Video Variational Autoencoders (Video VAEs) to compress intricate video data into a compact latent space. However, as LVDM training scales, the computational overhead of Video VAEs becomes a critical bottleneck, particularly for encoding high-resolution videos. To address this, we propose LeanVAE, a novel and ultra-efficient Video VAE framework that introduces two key innovations: (1) a lightweight architecture based on a Neighborhood-Aware Feedforward (NAF) module and non-overlapping pa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.14325","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/2503.14325/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":"2503.14325","created_at":"2026-07-05T10:34:00.227336+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.14325v1","created_at":"2026-07-05T10:34:00.227336+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.14325","created_at":"2026-07-05T10:34:00.227336+00:00"},{"alias_kind":"pith_short_12","alias_value":"XHYEK46B7WBN","created_at":"2026-07-05T10:34:00.227336+00:00"},{"alias_kind":"pith_short_16","alias_value":"XHYEK46B7WBNPQA4","created_at":"2026-07-05T10:34:00.227336+00:00"},{"alias_kind":"pith_short_8","alias_value":"XHYEK46B","created_at":"2026-07-05T10:34:00.227336+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01701","citing_title":"Arachne: Orchestrating Cascades for Efficient Text-to-Video Model Training","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30893","citing_title":"Foundation VAEs for 3D CT Reconstruction, Augmentation, and Generation","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17248","citing_title":"Image-to-Video Diffusion: From Foundations to Open Frontiers","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15911","citing_title":"Efficient Video Diffusion Models: Advancements and Challenges","ref_index":246,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7","json":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7.json","graph_json":"https://pith.science/api/pith-number/XHYEK46B7WBNPQA4HJBNSS2RT7/graph.json","events_json":"https://pith.science/api/pith-number/XHYEK46B7WBNPQA4HJBNSS2RT7/events.json","paper":"https://pith.science/paper/XHYEK46B"},"agent_actions":{"view_html":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7","download_json":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7.json","view_paper":"https://pith.science/paper/XHYEK46B","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.14325&json=true","fetch_graph":"https://pith.science/api/pith-number/XHYEK46B7WBNPQA4HJBNSS2RT7/graph.json","fetch_events":"https://pith.science/api/pith-number/XHYEK46B7WBNPQA4HJBNSS2RT7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7/action/storage_attestation","attest_author":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7/action/author_attestation","sign_citation":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7/action/citation_signature","submit_replication":"https://pith.science/pith/XHYEK46B7WBNPQA4HJBNSS2RT7/action/replication_record"}},"created_at":"2026-07-05T10:34:00.227336+00:00","updated_at":"2026-07-05T10:34:00.227336+00:00"}