{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:ZEM5FB3PQBSLZAL3JURK3NIWXY","short_pith_number":"pith:ZEM5FB3P","schema_version":"1.0","canonical_sha256":"c919d2876f8064bc817b4d22adb516be3fd1aed40e6c9897bcbd098ea14ce8ed","source":{"kind":"arxiv","id":"2107.00641","version":1},"attestation_state":"computed","paper":{"title":"Focal Self-attention for Local-Global Interactions in Vision Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Bin Xiao, Chunyuan Li, Jianfeng Gao, Jianwei Yang, Lu Yuan, Pengchuan Zhang, Xiyang Dai","submitted_at":"2021-07-01T17:56:09Z","abstract_excerpt":"Recently, Vision Transformer and its variants have shown great promise on various computer vision tasks. The ability of capturing short- and long-range visual dependencies through self-attention is arguably the main source for the success. But it also brings challenges due to quadratic computational overhead, especially for the high-resolution vision tasks (e.g., object detection). In this paper, we present focal self-attention, a new mechanism that incorporates both fine-grained local and coarse-grained global interactions. Using this new mechanism, each token attends the closest surrounding "},"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":"2107.00641","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-07-01T17:56:09Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"0e3b988c4665b96319683e50d594ec19be166704da5a62a15bd10830403cbc82","abstract_canon_sha256":"1377c996f95e7de14ab50fa72613475e423eeab494cc60f8f6957337c547bb25"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:54:26.693397Z","signature_b64":"4JIxISFB3iKZOSmYmEhb4rP9illBJfYGh0vD3f8783CTss8il91zAQ6lDbWjQZzGTD5wYN5UZv+jxnuPHvmaAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c919d2876f8064bc817b4d22adb516be3fd1aed40e6c9897bcbd098ea14ce8ed","last_reissued_at":"2026-07-05T02:54:26.692960Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:54:26.692960Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Focal Self-attention for Local-Global Interactions in Vision Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Bin Xiao, Chunyuan Li, Jianfeng Gao, Jianwei Yang, Lu Yuan, Pengchuan Zhang, Xiyang Dai","submitted_at":"2021-07-01T17:56:09Z","abstract_excerpt":"Recently, Vision Transformer and its variants have shown great promise on various computer vision tasks. The ability of capturing short- and long-range visual dependencies through self-attention is arguably the main source for the success. But it also brings challenges due to quadratic computational overhead, especially for the high-resolution vision tasks (e.g., object detection). In this paper, we present focal self-attention, a new mechanism that incorporates both fine-grained local and coarse-grained global interactions. Using this new mechanism, each token attends the closest surrounding "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2107.00641","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/2107.00641/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":"2107.00641","created_at":"2026-07-05T02:54:26.693026+00:00"},{"alias_kind":"arxiv_version","alias_value":"2107.00641v1","created_at":"2026-07-05T02:54:26.693026+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2107.00641","created_at":"2026-07-05T02:54:26.693026+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZEM5FB3PQBSL","created_at":"2026-07-05T02:54:26.693026+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZEM5FB3PQBSLZAL3","created_at":"2026-07-05T02:54:26.693026+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZEM5FB3P","created_at":"2026-07-05T02:54:26.693026+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2401.10166","citing_title":"VMamba: Visual State Space Model","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2111.11432","citing_title":"Florence: A New Foundation Model for Computer Vision","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2603.26357","citing_title":"MPDiT: Multi-Patch Global-to-Local Transformer Architecture For Efficient Flow Matching and Diffusion Model","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11300","citing_title":"Can Graphs Help Vision SSMs See Better?","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2401.09417","citing_title":"Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model","ref_index":78,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY","json":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY.json","graph_json":"https://pith.science/api/pith-number/ZEM5FB3PQBSLZAL3JURK3NIWXY/graph.json","events_json":"https://pith.science/api/pith-number/ZEM5FB3PQBSLZAL3JURK3NIWXY/events.json","paper":"https://pith.science/paper/ZEM5FB3P"},"agent_actions":{"view_html":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY","download_json":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY.json","view_paper":"https://pith.science/paper/ZEM5FB3P","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2107.00641&json=true","fetch_graph":"https://pith.science/api/pith-number/ZEM5FB3PQBSLZAL3JURK3NIWXY/graph.json","fetch_events":"https://pith.science/api/pith-number/ZEM5FB3PQBSLZAL3JURK3NIWXY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY/action/storage_attestation","attest_author":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY/action/author_attestation","sign_citation":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY/action/citation_signature","submit_replication":"https://pith.science/pith/ZEM5FB3PQBSLZAL3JURK3NIWXY/action/replication_record"}},"created_at":"2026-07-05T02:54:26.693026+00:00","updated_at":"2026-07-05T02:54:26.693026+00:00"}