{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:XVBR3LXYJRLEIRZ5EXWD32SS5C","short_pith_number":"pith:XVBR3LXY","schema_version":"1.0","canonical_sha256":"bd431daef84c5644473d25ec3dea52e896047632c5553fe785997f0d31ad6add","source":{"kind":"arxiv","id":"2211.10526","version":5},"attestation_state":"computed","paper":{"title":"Castling-ViT: Compressing Self-Attention via Switching Towards Linear-Angular Attention at Vision Transformer Inference","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bichen Wu, Haoqi Fan, Haoran You, Peizhao Zhang, Peter Vajda, Xiaoliang Dai, Yingyan Celine Lin, Yunyang Xiong","submitted_at":"2022-11-18T22:49:04Z","abstract_excerpt":"Vision Transformers (ViTs) have shown impressive performance but still require a high computation cost as compared to convolutional neural networks (CNNs), one reason is that ViTs' attention measures global similarities and thus has a quadratic complexity with the number of input tokens. Existing efficient ViTs adopt local attention (e.g., Swin) or linear attention (e.g., Performer), which sacrifice ViTs' capabilities of capturing either global or local context. In this work, we ask an important research question: Can ViTs learn both global and local context while being more efficient during i"},"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":"2211.10526","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2022-11-18T22:49:04Z","cross_cats_sorted":[],"title_canon_sha256":"fbcdeb3297970ed70dbb8f69dd99bee94e9a69485487060e6883cc1062c0d3e4","abstract_canon_sha256":"6033a9198dc875871300bf04bf8fc368db1017a95cac86c974e94b1305795e06"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:48:29.153891Z","signature_b64":"H5ODF5et25BJ3sWnJSWl0F+SoiSUJe9NBxVqBu2XVmxPcfxtj6wzwK1fuGOy12dIu+7Fjw1Sl6u22uJL8K+KCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bd431daef84c5644473d25ec3dea52e896047632c5553fe785997f0d31ad6add","last_reissued_at":"2026-07-05T08:48:29.153537Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:48:29.153537Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Castling-ViT: Compressing Self-Attention via Switching Towards Linear-Angular Attention at Vision Transformer Inference","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bichen Wu, Haoqi Fan, Haoran You, Peizhao Zhang, Peter Vajda, Xiaoliang Dai, Yingyan Celine Lin, Yunyang Xiong","submitted_at":"2022-11-18T22:49:04Z","abstract_excerpt":"Vision Transformers (ViTs) have shown impressive performance but still require a high computation cost as compared to convolutional neural networks (CNNs), one reason is that ViTs' attention measures global similarities and thus has a quadratic complexity with the number of input tokens. Existing efficient ViTs adopt local attention (e.g., Swin) or linear attention (e.g., Performer), which sacrifice ViTs' capabilities of capturing either global or local context. In this work, we ask an important research question: Can ViTs learn both global and local context while being more efficient during i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.10526","kind":"arxiv","version":5},"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/2211.10526/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":"2211.10526","created_at":"2026-07-05T08:48:29.153598+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.10526v5","created_at":"2026-07-05T08:48:29.153598+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.10526","created_at":"2026-07-05T08:48:29.153598+00:00"},{"alias_kind":"pith_short_12","alias_value":"XVBR3LXYJRLE","created_at":"2026-07-05T08:48:29.153598+00:00"},{"alias_kind":"pith_short_16","alias_value":"XVBR3LXYJRLEIRZ5","created_at":"2026-07-05T08:48:29.153598+00:00"},{"alias_kind":"pith_short_8","alias_value":"XVBR3LXY","created_at":"2026-07-05T08:48:29.153598+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/XVBR3LXYJRLEIRZ5EXWD32SS5C","json":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C.json","graph_json":"https://pith.science/api/pith-number/XVBR3LXYJRLEIRZ5EXWD32SS5C/graph.json","events_json":"https://pith.science/api/pith-number/XVBR3LXYJRLEIRZ5EXWD32SS5C/events.json","paper":"https://pith.science/paper/XVBR3LXY"},"agent_actions":{"view_html":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C","download_json":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C.json","view_paper":"https://pith.science/paper/XVBR3LXY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.10526&json=true","fetch_graph":"https://pith.science/api/pith-number/XVBR3LXYJRLEIRZ5EXWD32SS5C/graph.json","fetch_events":"https://pith.science/api/pith-number/XVBR3LXYJRLEIRZ5EXWD32SS5C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C/action/storage_attestation","attest_author":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C/action/author_attestation","sign_citation":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C/action/citation_signature","submit_replication":"https://pith.science/pith/XVBR3LXYJRLEIRZ5EXWD32SS5C/action/replication_record"}},"created_at":"2026-07-05T08:48:29.153598+00:00","updated_at":"2026-07-05T08:48:29.153598+00:00"}