{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:Q5GIJB5ZPBI2Q3Q2G3P44UVK5A","short_pith_number":"pith:Q5GIJB5Z","schema_version":"1.0","canonical_sha256":"874c8487b97851a86e1a36dfce52aae8169d2d7f8cee6e6e45435ead56131236","source":{"kind":"arxiv","id":"2112.05359","version":1},"attestation_state":"computed","paper":{"title":"Sketching as a Tool for Understanding and Accelerating Self-attention for Long Sequences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Di Jin, Dilek Hakkani-Tur, Heng Ji, Qi Zeng, Yifan Chen, Yun Yang","submitted_at":"2021-12-10T06:58:05Z","abstract_excerpt":"Transformer-based models are not efficient in processing long sequences due to the quadratic space and time complexity of the self-attention modules. To address this limitation, Linformer and Informer are proposed to reduce the quadratic complexity to linear (modulo logarithmic factors) via low-dimensional projection and row selection respectively. These two models are intrinsically connected, and to understand their connection, we introduce a theoretical framework of matrix sketching. Based on the theoretical analysis, we propose Skeinformer to accelerate self-attention and further improve th"},"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":"2112.05359","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-12-10T06:58:05Z","cross_cats_sorted":["cs.CL","stat.ML"],"title_canon_sha256":"5124dd19d572bc75b4f2dcefa32a7cb5eddf5b944c66beda729fddc3c9382f78","abstract_canon_sha256":"e1036e21284df549d9d4addd2815e90e4eacb17ebe084d18febc652765a94f94"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:39:40.390914Z","signature_b64":"HW4K6Gd0wK6NCvSCOFRmhUSzq2hp2MmOD0jMK7sDvA6tygDDyT5iN/YSuoh5FCW+xr9EV1denKiJq/E+sI9rDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"874c8487b97851a86e1a36dfce52aae8169d2d7f8cee6e6e45435ead56131236","last_reissued_at":"2026-07-05T03:39:40.390463Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:39:40.390463Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sketching as a Tool for Understanding and Accelerating Self-attention for Long Sequences","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Di Jin, Dilek Hakkani-Tur, Heng Ji, Qi Zeng, Yifan Chen, Yun Yang","submitted_at":"2021-12-10T06:58:05Z","abstract_excerpt":"Transformer-based models are not efficient in processing long sequences due to the quadratic space and time complexity of the self-attention modules. To address this limitation, Linformer and Informer are proposed to reduce the quadratic complexity to linear (modulo logarithmic factors) via low-dimensional projection and row selection respectively. These two models are intrinsically connected, and to understand their connection, we introduce a theoretical framework of matrix sketching. Based on the theoretical analysis, we propose Skeinformer to accelerate self-attention and further improve th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.05359","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/2112.05359/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":"2112.05359","created_at":"2026-07-05T03:39:40.390520+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.05359v1","created_at":"2026-07-05T03:39:40.390520+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.05359","created_at":"2026-07-05T03:39:40.390520+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q5GIJB5ZPBI2","created_at":"2026-07-05T03:39:40.390520+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q5GIJB5ZPBI2Q3Q2","created_at":"2026-07-05T03:39:40.390520+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q5GIJB5Z","created_at":"2026-07-05T03:39:40.390520+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/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A","json":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A.json","graph_json":"https://pith.science/api/pith-number/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/graph.json","events_json":"https://pith.science/api/pith-number/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/events.json","paper":"https://pith.science/paper/Q5GIJB5Z"},"agent_actions":{"view_html":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A","download_json":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A.json","view_paper":"https://pith.science/paper/Q5GIJB5Z","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.05359&json=true","fetch_graph":"https://pith.science/api/pith-number/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/graph.json","fetch_events":"https://pith.science/api/pith-number/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/action/storage_attestation","attest_author":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/action/author_attestation","sign_citation":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/action/citation_signature","submit_replication":"https://pith.science/pith/Q5GIJB5ZPBI2Q3Q2G3P44UVK5A/action/replication_record"}},"created_at":"2026-07-05T03:39:40.390520+00:00","updated_at":"2026-07-05T03:39:40.390520+00:00"}