{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN","short_pith_number":"pith:TJZ4X5Z3","canonical_record":{"source":{"id":"2006.04768","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-08T17:37:52Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"b5efd79fc27856c2c0e08c6b6be2d77bca3e4535ab2b0f02c37293802bc2e9ef","abstract_canon_sha256":"0556f2b940c7c5bbbba7a010eb8e5581eed12c0783236e58aa8dc6b3c4621650"},"schema_version":"1.0"},"canonical_sha256":"9a73cbf73b43a1e353a573e1bb921d8b536f35c1e1325dc4d516c765a5766f55","source":{"kind":"arxiv","id":"2006.04768","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2006.04768","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"arxiv_version","alias_value":"2006.04768v3","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.04768","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"pith_short_12","alias_value":"TJZ4X5Z3IOQ6","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"pith_short_16","alias_value":"TJZ4X5Z3IOQ6GU5F","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"pith_short_8","alias_value":"TJZ4X5Z3","created_at":"2026-07-05T01:09:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN","target":"record","payload":{"canonical_record":{"source":{"id":"2006.04768","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-08T17:37:52Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"b5efd79fc27856c2c0e08c6b6be2d77bca3e4535ab2b0f02c37293802bc2e9ef","abstract_canon_sha256":"0556f2b940c7c5bbbba7a010eb8e5581eed12c0783236e58aa8dc6b3c4621650"},"schema_version":"1.0"},"canonical_sha256":"9a73cbf73b43a1e353a573e1bb921d8b536f35c1e1325dc4d516c765a5766f55","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:09:59.491288Z","signature_b64":"syKotT7VnbzUtMf21B7xaSXdUQagMp6XavZNxxnQfYkcjI/IwWUiY0kUhIfyCYPKkk9BNaj0qOz7D8TYMTbmDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a73cbf73b43a1e353a573e1bb921d8b536f35c1e1325dc4d516c765a5766f55","last_reissued_at":"2026-07-05T01:09:59.490874Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:09:59.490874Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2006.04768","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T01:09:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rbYRmeowxKQB/KPQFkPNgYeMdRdYsE36myNShNh+IVbwN8mKcNCGyOF5c2p/G6u0AfRNfEpJGH1ly/Gl1OPyAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T23:44:18.939563Z"},"content_sha256":"2687121b89f65394a598236dc01ca44825bfe999b10081cf4e9ff79e52f10b5a","schema_version":"1.0","event_id":"sha256:2687121b89f65394a598236dc01ca44825bfe999b10081cf4e9ff79e52f10b5a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Linformer: Self-Attention with Linear Complexity","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Self-attention in transformers can be approximated by a low-rank matrix to reduce complexity to linear in sequence length.","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Belinda Z. Li, Han Fang, Hao Ma, Madian Khabsa, Sinong Wang","submitted_at":"2020-06-08T17:37:52Z","abstract_excerpt":"Large transformer models have shown extraordinary success in achieving state-of-the-art results in many natural language processing applications. However, training and deploying these models can be prohibitively costly for long sequences, as the standard self-attention mechanism of the Transformer uses $O(n^2)$ time and space with respect to sequence length. In this paper, we demonstrate that the self-attention mechanism can be approximated by a low-rank matrix. We further exploit this finding to propose a new self-attention mechanism, which reduces the overall self-attention complexity from $"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We demonstrate that the self-attention mechanism can be approximated by a low-rank matrix... The resulting linear transformer, the Linformer, performs on par with standard Transformer models, while being much more memory- and time-efficient.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the low-rank projection matrices learned or fixed during training preserve enough information for downstream tasks across the full range of sequence lengths and domains the model will encounter.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Linformer approximates self-attention with a low-rank projection to achieve O(n) time and space complexity while matching Transformer accuracy on standard NLP tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Self-attention in transformers can be approximated by a low-rank matrix to reduce complexity to linear in sequence length.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a487fe0efcf41a99445598c202731d9fbab80e318271895095e67dae94cf6208"},"source":{"id":"2006.04768","kind":"arxiv","version":3},"verdict":{"id":"0a08b9c0-8e77-493b-8931-242caa882ea7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T00:29:37.213350Z","strongest_claim":"We demonstrate that the self-attention mechanism can be approximated by a low-rank matrix... The resulting linear transformer, the Linformer, performs on par with standard Transformer models, while being much more memory- and time-efficient.","one_line_summary":"Linformer approximates self-attention with a low-rank projection to achieve O(n) time and space complexity while matching Transformer accuracy on standard NLP tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the low-rank projection matrices learned or fixed during training preserve enough information for downstream tasks across the full range of sequence lengths and domains the model will encounter.","pith_extraction_headline":"Self-attention in transformers can be approximated by a low-rank matrix to reduce complexity to linear in sequence length."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2006.04768/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":19,"sample":[{"doi":"","year":2004,"title":"Longformer: The Long-Document Transformer","work_id":"abea7a44-6668-4de7-aab6-f53a6e5aa088","ref_index":1,"cited_arxiv_id":"2004.05150","is_internal_anchor":true},{"doi":"","year":2005,"title":"Language Models are Few-Shot Learners","work_id":"214732c0-2edd-44a0-af9e-28184a2b8279","ref_index":2,"cited_arxiv_id":"2005.14165","is_internal_anchor":true},{"doi":"","year":null,"title":"Training Deep Nets with Sublinear Memory Cost","work_id":"f2c5c287-a500-40e4-a136-e7e3172db1d7","ref_index":3,"cited_arxiv_id":"1604.06174","is_internal_anchor":true},{"doi":"","year":1904,"title":"Generating Long Sequences with Sparse Transformers","work_id":"c5b81688-45ee-4a9a-b095-e6290f45cb6c","ref_index":4,"cited_arxiv_id":"1904.10509","is_internal_anchor":true},{"doi":"","year":2019,"title":"Bert: Pre-training of deep bidirectional transformers for language understanding","work_id":"693b70ad-3022-4615-938e-7752341ec181","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":19,"snapshot_sha256":"d7ab771ae1b9b3bcd851ffe57f3a2ee32c9ae27ca7672a6d6b9c0d536b970ec6","internal_anchors":11},"formal_canon":{"evidence_count":3,"snapshot_sha256":"2b22cdf79b79855b10388867a1a43a9d35d290d2b3eb673140f21aa4e921a432"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"0a08b9c0-8e77-493b-8931-242caa882ea7"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T01:09:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fUjALVtGv+mwwmQ6aDWWOAc1cp28YjY8T8gUZzvk3iJoQGVpmOfRdv/0/BhD0jLwUiYIjYRr9YL71Pl3mhpbCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T23:44:18.940424Z"},"content_sha256":"16501f2b7d51392fa6fb58a318e0df51c718febbdfed717f7761a56136b5b6e4","schema_version":"1.0","event_id":"sha256:16501f2b7d51392fa6fb58a318e0df51c718febbdfed717f7761a56136b5b6e4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN/bundle.json","state_url":"https://pith.science/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-15T23:44:18Z","links":{"resolver":"https://pith.science/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN","bundle":"https://pith.science/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN/bundle.json","state":"https://pith.science/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:TJZ4X5Z3IOQ6GU5FOPQ3XEQ5RN","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"0556f2b940c7c5bbbba7a010eb8e5581eed12c0783236e58aa8dc6b3c4621650","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-08T17:37:52Z","title_canon_sha256":"b5efd79fc27856c2c0e08c6b6be2d77bca3e4535ab2b0f02c37293802bc2e9ef"},"schema_version":"1.0","source":{"id":"2006.04768","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2006.04768","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"arxiv_version","alias_value":"2006.04768v3","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2006.04768","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"pith_short_12","alias_value":"TJZ4X5Z3IOQ6","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"pith_short_16","alias_value":"TJZ4X5Z3IOQ6GU5F","created_at":"2026-07-05T01:09:59Z"},{"alias_kind":"pith_short_8","alias_value":"TJZ4X5Z3","created_at":"2026-07-05T01:09:59Z"}],"graph_snapshots":[{"event_id":"sha256:16501f2b7d51392fa6fb58a318e0df51c718febbdfed717f7761a56136b5b6e4","target":"graph","created_at":"2026-07-05T01:09:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"We demonstrate that the self-attention mechanism can be approximated by a low-rank matrix... The resulting linear transformer, the Linformer, performs on par with standard Transformer models, while being much more memory- and time-efficient."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the low-rank projection matrices learned or fixed during training preserve enough information for downstream tasks across the full range of sequence lengths and domains the model will encounter."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Linformer approximates self-attention with a low-rank projection to achieve O(n) time and space complexity while matching Transformer accuracy on standard NLP tasks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Self-attention in transformers can be approximated by a low-rank matrix to reduce complexity to linear in sequence length."}],"snapshot_sha256":"a487fe0efcf41a99445598c202731d9fbab80e318271895095e67dae94cf6208"},"formal_canon":{"evidence_count":3,"snapshot_sha256":"2b22cdf79b79855b10388867a1a43a9d35d290d2b3eb673140f21aa4e921a432"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2006.04768/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large transformer models have shown extraordinary success in achieving state-of-the-art results in many natural language processing applications. However, training and deploying these models can be prohibitively costly for long sequences, as the standard self-attention mechanism of the Transformer uses $O(n^2)$ time and space with respect to sequence length. In this paper, we demonstrate that the self-attention mechanism can be approximated by a low-rank matrix. We further exploit this finding to propose a new self-attention mechanism, which reduces the overall self-attention complexity from $","authors_text":"Belinda Z. Li, Han Fang, Hao Ma, Madian Khabsa, Sinong Wang","cross_cats":["stat.ML"],"headline":"Self-attention in transformers can be approximated by a low-rank matrix to reduce complexity to linear in sequence length.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-08T17:37:52Z","title":"Linformer: Self-Attention with Linear Complexity"},"references":{"count":19,"internal_anchors":11,"resolved_work":19,"sample":[{"cited_arxiv_id":"2004.05150","doi":"","is_internal_anchor":true,"ref_index":1,"title":"Longformer: The Long-Document Transformer","work_id":"abea7a44-6668-4de7-aab6-f53a6e5aa088","year":2004},{"cited_arxiv_id":"2005.14165","doi":"","is_internal_anchor":true,"ref_index":2,"title":"Language Models are Few-Shot Learners","work_id":"214732c0-2edd-44a0-af9e-28184a2b8279","year":2005},{"cited_arxiv_id":"1604.06174","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Training Deep Nets with Sublinear Memory Cost","work_id":"f2c5c287-a500-40e4-a136-e7e3172db1d7","year":null},{"cited_arxiv_id":"1904.10509","doi":"","is_internal_anchor":true,"ref_index":4,"title":"Generating Long Sequences with Sparse Transformers","work_id":"c5b81688-45ee-4a9a-b095-e6290f45cb6c","year":1904},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Bert: Pre-training of deep bidirectional transformers for language understanding","work_id":"693b70ad-3022-4615-938e-7752341ec181","year":2019}],"snapshot_sha256":"d7ab771ae1b9b3bcd851ffe57f3a2ee32c9ae27ca7672a6d6b9c0d536b970ec6"},"source":{"id":"2006.04768","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-12T00:29:37.213350Z","id":"0a08b9c0-8e77-493b-8931-242caa882ea7","model_set":{"reader":"grok-4.3"},"one_line_summary":"Linformer approximates self-attention with a low-rank projection to achieve O(n) time and space complexity while matching Transformer accuracy on standard NLP tasks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Self-attention in transformers can be approximated by a low-rank matrix to reduce complexity to linear in sequence length.","strongest_claim":"We demonstrate that the self-attention mechanism can be approximated by a low-rank matrix... The resulting linear transformer, the Linformer, performs on par with standard Transformer models, while being much more memory- and time-efficient.","weakest_assumption":"That the low-rank projection matrices learned or fixed during training preserve enough information for downstream tasks across the full range of sequence lengths and domains the model will encounter."}},"verdict_id":"0a08b9c0-8e77-493b-8931-242caa882ea7"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:2687121b89f65394a598236dc01ca44825bfe999b10081cf4e9ff79e52f10b5a","target":"record","created_at":"2026-07-05T01:09:59Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"0556f2b940c7c5bbbba7a010eb8e5581eed12c0783236e58aa8dc6b3c4621650","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-06-08T17:37:52Z","title_canon_sha256":"b5efd79fc27856c2c0e08c6b6be2d77bca3e4535ab2b0f02c37293802bc2e9ef"},"schema_version":"1.0","source":{"id":"2006.04768","kind":"arxiv","version":3}},"canonical_sha256":"9a73cbf73b43a1e353a573e1bb921d8b536f35c1e1325dc4d516c765a5766f55","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9a73cbf73b43a1e353a573e1bb921d8b536f35c1e1325dc4d516c765a5766f55","first_computed_at":"2026-07-05T01:09:59.490874Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T01:09:59.490874Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"syKotT7VnbzUtMf21B7xaSXdUQagMp6XavZNxxnQfYkcjI/IwWUiY0kUhIfyCYPKkk9BNaj0qOz7D8TYMTbmDg==","signature_status":"signed_v1","signed_at":"2026-07-05T01:09:59.491288Z","signed_message":"canonical_sha256_bytes"},"source_id":"2006.04768","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2687121b89f65394a598236dc01ca44825bfe999b10081cf4e9ff79e52f10b5a","sha256:16501f2b7d51392fa6fb58a318e0df51c718febbdfed717f7761a56136b5b6e4"],"state_sha256":"177cd85b29ac8d473ad7ad3a71d87bd6a21b1e358e631156c76a33f7fe732237"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AcM+QYdwFwdc0OziPk6EHbxr1mjYiZi8QyS84EmNN+znHNT2pzwdfAz/yzfknBcx1hgoql6PWBA88Y4dr9QqCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T23:44:18.945773Z","bundle_sha256":"ae1a66439caf0e8d022890dda3b0aee196b3ef7a6c36da6a75a89c753de5a5f2"}}