{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ACZUUHXXMZ3SS4DJVWRSZORODQ","short_pith_number":"pith:ACZUUHXX","schema_version":"1.0","canonical_sha256":"00b34a1ef76677297069ada32cba2e1c284c22103505cdc3e4d5134b58890941","source":{"kind":"arxiv","id":"2304.00287","version":2},"attestation_state":"computed","paper":{"title":"Vision Transformers with Mixed-Resolution Tokenization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Avram Golbert, Omer Levy, Tomer Ronen","submitted_at":"2023-04-01T10:39:46Z","abstract_excerpt":"Vision Transformer models process input images by dividing them into a spatially regular grid of equal-size patches. Conversely, Transformers were originally introduced over natural language sequences, where each token represents a subword - a chunk of raw data of arbitrary size. In this work, we apply this approach to Vision Transformers by introducing a novel image tokenization scheme, replacing the standard uniform grid with a mixed-resolution sequence of tokens, where each token represents a patch of arbitrary size. Using the Quadtree algorithm and a novel saliency scorer, we construct a p"},"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":"2304.00287","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-04-01T10:39:46Z","cross_cats_sorted":[],"title_canon_sha256":"6bfaf8d13740cc0c56215b3cb43f902d7d7b366419e8a192415271638acdba17","abstract_canon_sha256":"f72c27c0528771a6d26dd288717aa15659df79bc804ae5feb3eff227d5969013"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:04:55.230756Z","signature_b64":"pLlTIXfNktrwEPKLafYdg+yyeDjOevMJn8dsVmOYerjC4U6JBnT7Cqjb0/UDZNXF62Qy19qzY+uPrsOVN02nBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"00b34a1ef76677297069ada32cba2e1c284c22103505cdc3e4d5134b58890941","last_reissued_at":"2026-07-05T06:04:55.230369Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:04:55.230369Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Vision Transformers with Mixed-Resolution Tokenization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Avram Golbert, Omer Levy, Tomer Ronen","submitted_at":"2023-04-01T10:39:46Z","abstract_excerpt":"Vision Transformer models process input images by dividing them into a spatially regular grid of equal-size patches. Conversely, Transformers were originally introduced over natural language sequences, where each token represents a subword - a chunk of raw data of arbitrary size. In this work, we apply this approach to Vision Transformers by introducing a novel image tokenization scheme, replacing the standard uniform grid with a mixed-resolution sequence of tokens, where each token represents a patch of arbitrary size. Using the Quadtree algorithm and a novel saliency scorer, we construct a p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.00287","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/2304.00287/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":"2304.00287","created_at":"2026-07-05T06:04:55.230428+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.00287v2","created_at":"2026-07-05T06:04:55.230428+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.00287","created_at":"2026-07-05T06:04:55.230428+00:00"},{"alias_kind":"pith_short_12","alias_value":"ACZUUHXXMZ3S","created_at":"2026-07-05T06:04:55.230428+00:00"},{"alias_kind":"pith_short_16","alias_value":"ACZUUHXXMZ3SS4DJ","created_at":"2026-07-05T06:04:55.230428+00:00"},{"alias_kind":"pith_short_8","alias_value":"ACZUUHXX","created_at":"2026-07-05T06:04:55.230428+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.15190","citing_title":"SemToken: Semantic-Aware Tokenization for Efficient Long-Context Language Modeling","ref_index":27,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ","json":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ.json","graph_json":"https://pith.science/api/pith-number/ACZUUHXXMZ3SS4DJVWRSZORODQ/graph.json","events_json":"https://pith.science/api/pith-number/ACZUUHXXMZ3SS4DJVWRSZORODQ/events.json","paper":"https://pith.science/paper/ACZUUHXX"},"agent_actions":{"view_html":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ","download_json":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ.json","view_paper":"https://pith.science/paper/ACZUUHXX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.00287&json=true","fetch_graph":"https://pith.science/api/pith-number/ACZUUHXXMZ3SS4DJVWRSZORODQ/graph.json","fetch_events":"https://pith.science/api/pith-number/ACZUUHXXMZ3SS4DJVWRSZORODQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ/action/storage_attestation","attest_author":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ/action/author_attestation","sign_citation":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ/action/citation_signature","submit_replication":"https://pith.science/pith/ACZUUHXXMZ3SS4DJVWRSZORODQ/action/replication_record"}},"created_at":"2026-07-05T06:04:55.230428+00:00","updated_at":"2026-07-05T06:04:55.230428+00:00"}