{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:2HXAXU64JKM3KQTVSZB726RZEE","short_pith_number":"pith:2HXAXU64","schema_version":"1.0","canonical_sha256":"d1ee0bd3dc4a99b542759643fd7a392137ea26bf9407dda2fa67929f03e2370d","source":{"kind":"arxiv","id":"2412.06590","version":1},"attestation_state":"computed","paper":{"title":"Bridging the Divide: Reconsidering Softmax and Linear Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dongchen Han, Gao Huang, Jiwen Lu, Shiji Song, Xiu Li, Xuran Pan, Yifan Pu, Yizeng Han, Zhuofan Xia","submitted_at":"2024-12-09T15:44:22Z","abstract_excerpt":"Widely adopted in modern Vision Transformer designs, Softmax attention can effectively capture long-range visual information; however, it incurs excessive computational cost when dealing with high-resolution inputs. In contrast, linear attention naturally enjoys linear complexity and has great potential to scale up to higher-resolution images. Nonetheless, the unsatisfactory performance of linear attention greatly limits its practical application in various scenarios. In this paper, we take a step forward to close the gap between the linear and Softmax attention with novel theoretical analyses"},"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":"2412.06590","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-12-09T15:44:22Z","cross_cats_sorted":[],"title_canon_sha256":"41869e6f63de349478dffa13c226b95ac302d6c135c437c9202e6f04185673ca","abstract_canon_sha256":"6af0eb2ed7bdc5cc586f2302b2e511b28a35bfd96d199702e1996c76b759c599"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:46:34.645059Z","signature_b64":"CTxnEdj65Bqncf3o2tfqEwkYTChF3jRMWzWquucPgbfgOqqp6ndUDXALtLaLeVvVrqCmva0E7HfTLIODNMYTAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d1ee0bd3dc4a99b542759643fd7a392137ea26bf9407dda2fa67929f03e2370d","last_reissued_at":"2026-07-05T09:46:34.644528Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:46:34.644528Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bridging the Divide: Reconsidering Softmax and Linear Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dongchen Han, Gao Huang, Jiwen Lu, Shiji Song, Xiu Li, Xuran Pan, Yifan Pu, Yizeng Han, Zhuofan Xia","submitted_at":"2024-12-09T15:44:22Z","abstract_excerpt":"Widely adopted in modern Vision Transformer designs, Softmax attention can effectively capture long-range visual information; however, it incurs excessive computational cost when dealing with high-resolution inputs. In contrast, linear attention naturally enjoys linear complexity and has great potential to scale up to higher-resolution images. Nonetheless, the unsatisfactory performance of linear attention greatly limits its practical application in various scenarios. In this paper, we take a step forward to close the gap between the linear and Softmax attention with novel theoretical analyses"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.06590","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/2412.06590/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":"2412.06590","created_at":"2026-07-05T09:46:34.644593+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.06590v1","created_at":"2026-07-05T09:46:34.644593+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.06590","created_at":"2026-07-05T09:46:34.644593+00:00"},{"alias_kind":"pith_short_12","alias_value":"2HXAXU64JKM3","created_at":"2026-07-05T09:46:34.644593+00:00"},{"alias_kind":"pith_short_16","alias_value":"2HXAXU64JKM3KQTV","created_at":"2026-07-05T09:46:34.644593+00:00"},{"alias_kind":"pith_short_8","alias_value":"2HXAXU64","created_at":"2026-07-05T09:46:34.644593+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.24788","citing_title":"GLIDE: Guided Layerwise Hybrid Attention for Efficient LLM Inference","ref_index":23,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE","json":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE.json","graph_json":"https://pith.science/api/pith-number/2HXAXU64JKM3KQTVSZB726RZEE/graph.json","events_json":"https://pith.science/api/pith-number/2HXAXU64JKM3KQTVSZB726RZEE/events.json","paper":"https://pith.science/paper/2HXAXU64"},"agent_actions":{"view_html":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE","download_json":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE.json","view_paper":"https://pith.science/paper/2HXAXU64","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.06590&json=true","fetch_graph":"https://pith.science/api/pith-number/2HXAXU64JKM3KQTVSZB726RZEE/graph.json","fetch_events":"https://pith.science/api/pith-number/2HXAXU64JKM3KQTVSZB726RZEE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE/action/storage_attestation","attest_author":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE/action/author_attestation","sign_citation":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE/action/citation_signature","submit_replication":"https://pith.science/pith/2HXAXU64JKM3KQTVSZB726RZEE/action/replication_record"}},"created_at":"2026-07-05T09:46:34.644593+00:00","updated_at":"2026-07-05T09:46:34.644593+00:00"}