{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5ECMBKMCLBBFH4AL6PMZJUARYG","short_pith_number":"pith:5ECMBKMC","schema_version":"1.0","canonical_sha256":"e904c0a982584253f00bf3d994d011c18e8c8c979e07f45262a1edca06899b41","source":{"kind":"arxiv","id":"2405.14477","version":2},"attestation_state":"computed","paper":{"title":"LiteVAE: Lightweight and Efficient Variational Autoencoders for Latent Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Derek Bradley, Jakob Buhmann, Otmar Hilliges, Romann M. Weber, Seyedmorteza Sadat","submitted_at":"2024-05-23T12:06:00Z","abstract_excerpt":"Advances in latent diffusion models (LDMs) have revolutionized high-resolution image generation, but the design space of the autoencoder that is central to these systems remains underexplored. In this paper, we introduce LiteVAE, a new autoencoder design for LDMs, which leverages the 2D discrete wavelet transform to enhance scalability and computational efficiency over standard variational autoencoders (VAEs) with no sacrifice in output quality. We investigate the training methodologies and the decoder architecture of LiteVAE and propose several enhancements that improve the training dynamics "},"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":"2405.14477","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-23T12:06:00Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"26ec0ffea9b09c925ea7944d37aff2498dd8cbaa911583b71d9afb99d4972cfa","abstract_canon_sha256":"7d23c85b010d092799b19617ae8eeb8f3f25b5432c0b38fb4a8269ffd27e822f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:03:25.342308Z","signature_b64":"UkzySW/hg2DwYyuHrNwB+ncDrzvSxW3NFCUCN+11dJ4MUM3JzkJmRSSDmRj0bLR+nChl3AsA2ucyPyArtdFBBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e904c0a982584253f00bf3d994d011c18e8c8c979e07f45262a1edca06899b41","last_reissued_at":"2026-07-05T10:03:25.341815Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:03:25.341815Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LiteVAE: Lightweight and Efficient Variational Autoencoders for Latent Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Derek Bradley, Jakob Buhmann, Otmar Hilliges, Romann M. Weber, Seyedmorteza Sadat","submitted_at":"2024-05-23T12:06:00Z","abstract_excerpt":"Advances in latent diffusion models (LDMs) have revolutionized high-resolution image generation, but the design space of the autoencoder that is central to these systems remains underexplored. In this paper, we introduce LiteVAE, a new autoencoder design for LDMs, which leverages the 2D discrete wavelet transform to enhance scalability and computational efficiency over standard variational autoencoders (VAEs) with no sacrifice in output quality. We investigate the training methodologies and the decoder architecture of LiteVAE and propose several enhancements that improve the training dynamics "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.14477","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/2405.14477/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":"2405.14477","created_at":"2026-07-05T10:03:25.341879+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.14477v2","created_at":"2026-07-05T10:03:25.341879+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.14477","created_at":"2026-07-05T10:03:25.341879+00:00"},{"alias_kind":"pith_short_12","alias_value":"5ECMBKMCLBBF","created_at":"2026-07-05T10:03:25.341879+00:00"},{"alias_kind":"pith_short_16","alias_value":"5ECMBKMCLBBFH4AL","created_at":"2026-07-05T10:03:25.341879+00:00"},{"alias_kind":"pith_short_8","alias_value":"5ECMBKMC","created_at":"2026-07-05T10:03:25.341879+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2501.03575","citing_title":"Cosmos World Foundation Model Platform for Physical AI","ref_index":170,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG","json":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG.json","graph_json":"https://pith.science/api/pith-number/5ECMBKMCLBBFH4AL6PMZJUARYG/graph.json","events_json":"https://pith.science/api/pith-number/5ECMBKMCLBBFH4AL6PMZJUARYG/events.json","paper":"https://pith.science/paper/5ECMBKMC"},"agent_actions":{"view_html":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG","download_json":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG.json","view_paper":"https://pith.science/paper/5ECMBKMC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.14477&json=true","fetch_graph":"https://pith.science/api/pith-number/5ECMBKMCLBBFH4AL6PMZJUARYG/graph.json","fetch_events":"https://pith.science/api/pith-number/5ECMBKMCLBBFH4AL6PMZJUARYG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG/action/storage_attestation","attest_author":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG/action/author_attestation","sign_citation":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG/action/citation_signature","submit_replication":"https://pith.science/pith/5ECMBKMCLBBFH4AL6PMZJUARYG/action/replication_record"}},"created_at":"2026-07-05T10:03:25.341879+00:00","updated_at":"2026-07-05T10:03:25.341879+00:00"}