{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:E2XB7VCFIQX2X2QJ6OB4TH7ZH5","short_pith_number":"pith:E2XB7VCF","schema_version":"1.0","canonical_sha256":"26ae1fd445442fabea09f383c99ff93f64c52b3626f07c20ca8173e852fad66c","source":{"kind":"arxiv","id":"2410.13925","version":1},"attestation_state":"computed","paper":{"title":"FiTv2: Scalable and Improved Flexible Vision Transformer for Diffusion Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"and Lei Bai, Cai Zhou, Di Huang, Wanli Ouyang, Zeyu Lu, Zidong Wang","submitted_at":"2024-10-17T15:51:49Z","abstract_excerpt":"\\textit{Nature is infinitely resolution-free}. In the context of this reality, existing diffusion models, such as Diffusion Transformers, often face challenges when processing image resolutions outside of their trained domain. To address this limitation, we conceptualize images as sequences of tokens with dynamic sizes, rather than traditional methods that perceive images as fixed-resolution grids. This perspective enables a flexible training strategy that seamlessly accommodates various aspect ratios during both training and inference, thus promoting resolution generalization and eliminating "},"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":"2410.13925","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-17T15:51:49Z","cross_cats_sorted":[],"title_canon_sha256":"eb48c512019de8e79e2a22224cd60ae1ed3a45348a08be066fd7b6b62170ec6d","abstract_canon_sha256":"e508ef1309684c7d555d243adf59ab382c88485dc43e063e25451a88c56195b6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:22:24.663468Z","signature_b64":"i8l9vr8+n0XVkpYRDXTnV+Mnl9I9yzYqPExMZEiItaBd3R1WJwsTkWiEyCBM5z39CCqDZ2nvcalFhV9KVYueDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"26ae1fd445442fabea09f383c99ff93f64c52b3626f07c20ca8173e852fad66c","last_reissued_at":"2026-07-05T09:22:24.662992Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:22:24.662992Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FiTv2: Scalable and Improved Flexible Vision Transformer for Diffusion Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"and Lei Bai, Cai Zhou, Di Huang, Wanli Ouyang, Zeyu Lu, Zidong Wang","submitted_at":"2024-10-17T15:51:49Z","abstract_excerpt":"\\textit{Nature is infinitely resolution-free}. In the context of this reality, existing diffusion models, such as Diffusion Transformers, often face challenges when processing image resolutions outside of their trained domain. To address this limitation, we conceptualize images as sequences of tokens with dynamic sizes, rather than traditional methods that perceive images as fixed-resolution grids. This perspective enables a flexible training strategy that seamlessly accommodates various aspect ratios during both training and inference, thus promoting resolution generalization and eliminating "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13925","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/2410.13925/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":"2410.13925","created_at":"2026-07-05T09:22:24.663048+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13925v1","created_at":"2026-07-05T09:22:24.663048+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13925","created_at":"2026-07-05T09:22:24.663048+00:00"},{"alias_kind":"pith_short_12","alias_value":"E2XB7VCFIQX2","created_at":"2026-07-05T09:22:24.663048+00:00"},{"alias_kind":"pith_short_16","alias_value":"E2XB7VCFIQX2X2QJ","created_at":"2026-07-05T09:22:24.663048+00:00"},{"alias_kind":"pith_short_8","alias_value":"E2XB7VCF","created_at":"2026-07-05T09:22:24.663048+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2412.00131","citing_title":"Open-Sora Plan: Open-Source Large Video Generation Model","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15908","citing_title":"RaPD: Resolution-Agnostic Pixel Diffusion via Semantics-Enriched Implicit Representations","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24885","citing_title":"VibeToken: Scaling 1D Image Tokenizers and Autoregressive Models for Dynamic Resolution Generations","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07397","citing_title":"Data Warmup: Complexity-Aware Curricula for Efficient Diffusion Training","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5","json":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5.json","graph_json":"https://pith.science/api/pith-number/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/graph.json","events_json":"https://pith.science/api/pith-number/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/events.json","paper":"https://pith.science/paper/E2XB7VCF"},"agent_actions":{"view_html":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5","download_json":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5.json","view_paper":"https://pith.science/paper/E2XB7VCF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13925&json=true","fetch_graph":"https://pith.science/api/pith-number/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/graph.json","fetch_events":"https://pith.science/api/pith-number/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/action/storage_attestation","attest_author":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/action/author_attestation","sign_citation":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/action/citation_signature","submit_replication":"https://pith.science/pith/E2XB7VCFIQX2X2QJ6OB4TH7ZH5/action/replication_record"}},"created_at":"2026-07-05T09:22:24.663048+00:00","updated_at":"2026-07-05T09:22:24.663048+00:00"}