{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:BBAOAYA4Q2WWG2YAAEW2C5EASO","short_pith_number":"pith:BBAOAYA4","canonical_record":{"source":{"id":"2405.18616","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T21:45:46Z","cross_cats_sorted":[],"title_canon_sha256":"ce9d58303e5a3c86ffe0313700142cf3745d77c8bf57ded298c605390482ec5c","abstract_canon_sha256":"64491089d51b4c44c8799dbee81bbccead60d0b61f02ff25bec32d924270b50e"},"schema_version":"1.0"},"canonical_sha256":"0840e0601c86ad636b00012da1748093a3f6c9480269e7ba7910e04288444669","source":{"kind":"arxiv","id":"2405.18616","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.18616","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"arxiv_version","alias_value":"2405.18616v1","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.18616","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"pith_short_12","alias_value":"BBAOAYA4Q2WW","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"pith_short_16","alias_value":"BBAOAYA4Q2WWG2YA","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"pith_short_8","alias_value":"BBAOAYA4","created_at":"2026-07-05T08:24:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:BBAOAYA4Q2WWG2YAAEW2C5EASO","target":"record","payload":{"canonical_record":{"source":{"id":"2405.18616","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T21:45:46Z","cross_cats_sorted":[],"title_canon_sha256":"ce9d58303e5a3c86ffe0313700142cf3745d77c8bf57ded298c605390482ec5c","abstract_canon_sha256":"64491089d51b4c44c8799dbee81bbccead60d0b61f02ff25bec32d924270b50e"},"schema_version":"1.0"},"canonical_sha256":"0840e0601c86ad636b00012da1748093a3f6c9480269e7ba7910e04288444669","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:28.100839Z","signature_b64":"3/rgGSPysBktl6TcWuf4fkOq0FtHwk6WGtRP/2dbxjYsHfPW+9JopktY3BqZoCEnL0WORxicRCYc64Md1BciDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0840e0601c86ad636b00012da1748093a3f6c9480269e7ba7910e04288444669","last_reissued_at":"2026-07-05T08:24:28.100304Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:28.100304Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.18616","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:24:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QykjSWYSp+mUSCa1xhaljzarmgJ7xbwilLx+vt2vOUFCJKpn0B8h9qwqK+6N96gjAHQqdDy3k0chdHTqQ6wRBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T12:30:27.755521Z"},"content_sha256":"d6875bc17ea54792d88e12e4778cd28a8ae4af2f7a906f87387e1c26a0dda171","schema_version":"1.0","event_id":"sha256:d6875bc17ea54792d88e12e4778cd28a8ae4af2f7a906f87387e1c26a0dda171"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:BBAOAYA4Q2WWG2YAAEW2C5EASO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Wavelet-Based Image Tokenizer for Vision Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Radu Soricut, Zhenhai Zhu","submitted_at":"2024-05-28T21:45:46Z","abstract_excerpt":"Non-overlapping patch-wise convolution is the default image tokenizer for all state-of-the-art vision Transformer (ViT) models. Even though many ViT variants have been proposed to improve its efficiency and accuracy, little research on improving the image tokenizer itself has been reported in the literature. In this paper, we propose a new image tokenizer based on wavelet transformation. We show that ViT models with the new tokenizer achieve both higher training throughput and better top-1 precision for the ImageNet validation set. We present a theoretical analysis on why the proposed tokenize"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.18616","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/2405.18616/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:24:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kM0Ic+lp/GGKWVlObviEVaC6P3ExgUuLZOMJcnn4w8DT/RpAZtHzbWw6RJxZubUQb5iH6e80p8JLmktrhgO3BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T12:30:27.756012Z"},"content_sha256":"e47e7f2b1c771692c899e4adef2e0bf86d3106b28c0a4eb66afe2b0016149ade","schema_version":"1.0","event_id":"sha256:e47e7f2b1c771692c899e4adef2e0bf86d3106b28c0a4eb66afe2b0016149ade"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO/bundle.json","state_url":"https://pith.science/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-08T12:30:27Z","links":{"resolver":"https://pith.science/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO","bundle":"https://pith.science/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO/bundle.json","state":"https://pith.science/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BBAOAYA4Q2WWG2YAAEW2C5EASO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:BBAOAYA4Q2WWG2YAAEW2C5EASO","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"64491089d51b4c44c8799dbee81bbccead60d0b61f02ff25bec32d924270b50e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T21:45:46Z","title_canon_sha256":"ce9d58303e5a3c86ffe0313700142cf3745d77c8bf57ded298c605390482ec5c"},"schema_version":"1.0","source":{"id":"2405.18616","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.18616","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"arxiv_version","alias_value":"2405.18616v1","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.18616","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"pith_short_12","alias_value":"BBAOAYA4Q2WW","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"pith_short_16","alias_value":"BBAOAYA4Q2WWG2YA","created_at":"2026-07-05T08:24:28Z"},{"alias_kind":"pith_short_8","alias_value":"BBAOAYA4","created_at":"2026-07-05T08:24:28Z"}],"graph_snapshots":[{"event_id":"sha256:e47e7f2b1c771692c899e4adef2e0bf86d3106b28c0a4eb66afe2b0016149ade","target":"graph","created_at":"2026-07-05T08:24:28Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2405.18616/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Non-overlapping patch-wise convolution is the default image tokenizer for all state-of-the-art vision Transformer (ViT) models. Even though many ViT variants have been proposed to improve its efficiency and accuracy, little research on improving the image tokenizer itself has been reported in the literature. In this paper, we propose a new image tokenizer based on wavelet transformation. We show that ViT models with the new tokenizer achieve both higher training throughput and better top-1 precision for the ImageNet validation set. We present a theoretical analysis on why the proposed tokenize","authors_text":"Radu Soricut, Zhenhai Zhu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T21:45:46Z","title":"Wavelet-Based Image Tokenizer for Vision Transformers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.18616","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d6875bc17ea54792d88e12e4778cd28a8ae4af2f7a906f87387e1c26a0dda171","target":"record","created_at":"2026-07-05T08:24:28Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"64491089d51b4c44c8799dbee81bbccead60d0b61f02ff25bec32d924270b50e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T21:45:46Z","title_canon_sha256":"ce9d58303e5a3c86ffe0313700142cf3745d77c8bf57ded298c605390482ec5c"},"schema_version":"1.0","source":{"id":"2405.18616","kind":"arxiv","version":1}},"canonical_sha256":"0840e0601c86ad636b00012da1748093a3f6c9480269e7ba7910e04288444669","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0840e0601c86ad636b00012da1748093a3f6c9480269e7ba7910e04288444669","first_computed_at":"2026-07-05T08:24:28.100304Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:24:28.100304Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"3/rgGSPysBktl6TcWuf4fkOq0FtHwk6WGtRP/2dbxjYsHfPW+9JopktY3BqZoCEnL0WORxicRCYc64Md1BciDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:24:28.100839Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.18616","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d6875bc17ea54792d88e12e4778cd28a8ae4af2f7a906f87387e1c26a0dda171","sha256:e47e7f2b1c771692c899e4adef2e0bf86d3106b28c0a4eb66afe2b0016149ade"],"state_sha256":"0bf482d31109676a40db3e579736f0f352db2c19bd2b2a117aedaeb52492f510"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hyXx2xedqBnEsZC0tRHtv5FWcq62IdEjLiYKBco7CjCb6HiHXnJPRpo0w0G7LGwLtxmdVex3HiWdG57XKVfpDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T12:30:27.759677Z","bundle_sha256":"3e63c9ff3e453e9a34ade6d26574f910a90c0d7f1458098ea089aa46df8a4c41"}}