{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:JPLHMB3OWC7GUGQWXWSGBBQNM3","short_pith_number":"pith:JPLHMB3O","schema_version":"1.0","canonical_sha256":"4bd676076eb0be6a1a16bda460860d66dbbbcde220f794b7470d2f4236edf538","source":{"kind":"arxiv","id":"2605.26636","version":1},"attestation_state":"computed","paper":{"title":"JetViT: Efficient High-Resolution Vision Transformer with Post-Training Attention Search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Dongyun Zou, Han Cai, Hanrong Ye, Hongxu Yin, Junyu Chen, Qinhe Peng, Song Han, Wenkun He, Yao Lu, Yu Wang, Zhuoyang Zhang","submitted_at":"2026-05-26T07:17:27Z","abstract_excerpt":"We introduce JetViT, a novel family of hybrid-architecture Vision Transformer (ViT) models that match the accuracy of state-of-the-art full-attention vision foundation models while achieving substantially higher inference efficiency on high-resolution images. At the core of our approach is Post-Training Attention Search, a post-training acceleration framework that converts pre-trained full-attention ViTs into efficient hybrid-attention variants by identifying and replacing redundant full-attention blocks with linear or window-attention blocks. By inheriting the MLP and attention weights from t"},"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":"2605.26636","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-26T07:17:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"803b180610226bd676192c1e73a7ada8897616c8ae871959129f315b85ef775c","abstract_canon_sha256":"78d5da46c256b3e7b087c7a170e11d625880414483a82941df480fcae7b60c03"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:06:03.259341Z","signature_b64":"8ouT0lJP7tr3kR51MSCg5CXfSpH13uawETjf7UsCrGNz9jss/ytPA4JHXEuTydT90iAsCR++fooX4HGBg8IiBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4bd676076eb0be6a1a16bda460860d66dbbbcde220f794b7470d2f4236edf538","last_reissued_at":"2026-05-27T01:06:03.258649Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:06:03.258649Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"JetViT: Efficient High-Resolution Vision Transformer with Post-Training Attention Search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Dongyun Zou, Han Cai, Hanrong Ye, Hongxu Yin, Junyu Chen, Qinhe Peng, Song Han, Wenkun He, Yao Lu, Yu Wang, Zhuoyang Zhang","submitted_at":"2026-05-26T07:17:27Z","abstract_excerpt":"We introduce JetViT, a novel family of hybrid-architecture Vision Transformer (ViT) models that match the accuracy of state-of-the-art full-attention vision foundation models while achieving substantially higher inference efficiency on high-resolution images. At the core of our approach is Post-Training Attention Search, a post-training acceleration framework that converts pre-trained full-attention ViTs into efficient hybrid-attention variants by identifying and replacing redundant full-attention blocks with linear or window-attention blocks. By inheriting the MLP and attention weights from t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.26636","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/2605.26636/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":"2605.26636","created_at":"2026-05-27T01:06:03.258780+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.26636v1","created_at":"2026-05-27T01:06:03.258780+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.26636","created_at":"2026-05-27T01:06:03.258780+00:00"},{"alias_kind":"pith_short_12","alias_value":"JPLHMB3OWC7G","created_at":"2026-05-27T01:06:03.258780+00:00"},{"alias_kind":"pith_short_16","alias_value":"JPLHMB3OWC7GUGQW","created_at":"2026-05-27T01:06:03.258780+00:00"},{"alias_kind":"pith_short_8","alias_value":"JPLHMB3O","created_at":"2026-05-27T01:06:03.258780+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3","json":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3.json","graph_json":"https://pith.science/api/pith-number/JPLHMB3OWC7GUGQWXWSGBBQNM3/graph.json","events_json":"https://pith.science/api/pith-number/JPLHMB3OWC7GUGQWXWSGBBQNM3/events.json","paper":"https://pith.science/paper/JPLHMB3O"},"agent_actions":{"view_html":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3","download_json":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3.json","view_paper":"https://pith.science/paper/JPLHMB3O","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.26636&json=true","fetch_graph":"https://pith.science/api/pith-number/JPLHMB3OWC7GUGQWXWSGBBQNM3/graph.json","fetch_events":"https://pith.science/api/pith-number/JPLHMB3OWC7GUGQWXWSGBBQNM3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3/action/storage_attestation","attest_author":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3/action/author_attestation","sign_citation":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3/action/citation_signature","submit_replication":"https://pith.science/pith/JPLHMB3OWC7GUGQWXWSGBBQNM3/action/replication_record"}},"created_at":"2026-05-27T01:06:03.258780+00:00","updated_at":"2026-05-27T01:06:03.258780+00:00"}