{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:FMGYNXEW3PHQWIQKZKX76H5WJG","short_pith_number":"pith:FMGYNXEW","canonical_record":{"source":{"id":"2102.12702","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-02-25T06:18:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"59574fda5a180a407d1d2bec1037814cd1cb8948e7bfcfb067ace2ce85a8a83f","abstract_canon_sha256":"fe366ecc3f016ef1697a4b44cf5101093b9fab4f42294d2dc99a8ad5d5df9445"},"schema_version":"1.0"},"canonical_sha256":"2b0d86dc96dbcf0b220acaafff1fb6499669d3ebbdbe91cd07f1d4b01ae8d8e6","source":{"kind":"arxiv","id":"2102.12702","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2102.12702","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"arxiv_version","alias_value":"2102.12702v1","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.12702","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"pith_short_12","alias_value":"FMGYNXEW3PHQ","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"pith_short_16","alias_value":"FMGYNXEW3PHQWIQK","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"pith_short_8","alias_value":"FMGYNXEW","created_at":"2026-07-05T02:18:23Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:FMGYNXEW3PHQWIQKZKX76H5WJG","target":"record","payload":{"canonical_record":{"source":{"id":"2102.12702","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-02-25T06:18:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"59574fda5a180a407d1d2bec1037814cd1cb8948e7bfcfb067ace2ce85a8a83f","abstract_canon_sha256":"fe366ecc3f016ef1697a4b44cf5101093b9fab4f42294d2dc99a8ad5d5df9445"},"schema_version":"1.0"},"canonical_sha256":"2b0d86dc96dbcf0b220acaafff1fb6499669d3ebbdbe91cd07f1d4b01ae8d8e6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:18:23.053565Z","signature_b64":"byr8C9dipCnRLEpnV0LQyYpk9ciJ7B/FUSlpfmQ/ioX7ERiT1Zm79BCqiJ6kLy160ZBGluQR3yiNcGvUcHQ1Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b0d86dc96dbcf0b220acaafff1fb6499669d3ebbdbe91cd07f1d4b01ae8d8e6","last_reissued_at":"2026-07-05T02:18:23.053144Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:18:23.053144Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2102.12702","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T02:18:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3zXVMX/olv5Qb8EtxkNOBctaeVcHGjEMvKCbbNepSi5qk1xdW3OutQXQbYK5rLc930U5kLBKjjo7xtcdMHJOBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T00:01:50.567127Z"},"content_sha256":"9d977d788c8139de4e6603405377675552116f340c9ece3d38d8a64d6fb7bce8","schema_version":"1.0","event_id":"sha256:9d977d788c8139de4e6603405377675552116f340c9ece3d38d8a64d6fb7bce8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:FMGYNXEW3PHQWIQKZKX76H5WJG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LazyFormer: Self Attention with Lazy Update","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chengxuan Ying, Di He, Guolin Ke, Tie-Yan Liu","submitted_at":"2021-02-25T06:18:20Z","abstract_excerpt":"Improving the efficiency of Transformer-based language pre-training is an important task in NLP, especially for the self-attention module, which is computationally expensive. In this paper, we propose a simple but effective solution, called \\emph{LazyFormer}, which computes the self-attention distribution infrequently. LazyFormer composes of multiple lazy blocks, each of which contains multiple Transformer layers. In each lazy block, the self-attention distribution is only computed once in the first layer and then is reused in all upper layers. In this way, the cost of computation could be lar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.12702","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/2102.12702/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T02:18:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qmL1XVCYoBDTaNjz8Q7KEl/t1nH9sCEepWe2C202sFiDQiXbK3N3n3Y2pYRndWRWZ9FOHVBazQX97lOAy58VDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T00:01:50.567669Z"},"content_sha256":"f294a0db44376b2d17f1a0661d506b0848c7563b317708bc42d48ff599ebc29e","schema_version":"1.0","event_id":"sha256:f294a0db44376b2d17f1a0661d506b0848c7563b317708bc42d48ff599ebc29e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FMGYNXEW3PHQWIQKZKX76H5WJG/bundle.json","state_url":"https://pith.science/pith/FMGYNXEW3PHQWIQKZKX76H5WJG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FMGYNXEW3PHQWIQKZKX76H5WJG/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":"2026-08-04T00:01:50Z","links":{"resolver":"https://pith.science/pith/FMGYNXEW3PHQWIQKZKX76H5WJG","bundle":"https://pith.science/pith/FMGYNXEW3PHQWIQKZKX76H5WJG/bundle.json","state":"https://pith.science/pith/FMGYNXEW3PHQWIQKZKX76H5WJG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FMGYNXEW3PHQWIQKZKX76H5WJG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:FMGYNXEW3PHQWIQKZKX76H5WJG","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":"fe366ecc3f016ef1697a4b44cf5101093b9fab4f42294d2dc99a8ad5d5df9445","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-02-25T06:18:20Z","title_canon_sha256":"59574fda5a180a407d1d2bec1037814cd1cb8948e7bfcfb067ace2ce85a8a83f"},"schema_version":"1.0","source":{"id":"2102.12702","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2102.12702","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"arxiv_version","alias_value":"2102.12702v1","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.12702","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"pith_short_12","alias_value":"FMGYNXEW3PHQ","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"pith_short_16","alias_value":"FMGYNXEW3PHQWIQK","created_at":"2026-07-05T02:18:23Z"},{"alias_kind":"pith_short_8","alias_value":"FMGYNXEW","created_at":"2026-07-05T02:18:23Z"}],"graph_snapshots":[{"event_id":"sha256:f294a0db44376b2d17f1a0661d506b0848c7563b317708bc42d48ff599ebc29e","target":"graph","created_at":"2026-07-05T02:18:23Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2102.12702/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Improving the efficiency of Transformer-based language pre-training is an important task in NLP, especially for the self-attention module, which is computationally expensive. In this paper, we propose a simple but effective solution, called \\emph{LazyFormer}, which computes the self-attention distribution infrequently. LazyFormer composes of multiple lazy blocks, each of which contains multiple Transformer layers. In each lazy block, the self-attention distribution is only computed once in the first layer and then is reused in all upper layers. In this way, the cost of computation could be lar","authors_text":"Chengxuan Ying, Di He, Guolin Ke, Tie-Yan Liu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-02-25T06:18:20Z","title":"LazyFormer: Self Attention with Lazy Update"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.12702","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9d977d788c8139de4e6603405377675552116f340c9ece3d38d8a64d6fb7bce8","target":"record","created_at":"2026-07-05T02:18:23Z","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":"fe366ecc3f016ef1697a4b44cf5101093b9fab4f42294d2dc99a8ad5d5df9445","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-02-25T06:18:20Z","title_canon_sha256":"59574fda5a180a407d1d2bec1037814cd1cb8948e7bfcfb067ace2ce85a8a83f"},"schema_version":"1.0","source":{"id":"2102.12702","kind":"arxiv","version":1}},"canonical_sha256":"2b0d86dc96dbcf0b220acaafff1fb6499669d3ebbdbe91cd07f1d4b01ae8d8e6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2b0d86dc96dbcf0b220acaafff1fb6499669d3ebbdbe91cd07f1d4b01ae8d8e6","first_computed_at":"2026-07-05T02:18:23.053144Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:18:23.053144Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"byr8C9dipCnRLEpnV0LQyYpk9ciJ7B/FUSlpfmQ/ioX7ERiT1Zm79BCqiJ6kLy160ZBGluQR3yiNcGvUcHQ1Cw==","signature_status":"signed_v1","signed_at":"2026-07-05T02:18:23.053565Z","signed_message":"canonical_sha256_bytes"},"source_id":"2102.12702","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9d977d788c8139de4e6603405377675552116f340c9ece3d38d8a64d6fb7bce8","sha256:f294a0db44376b2d17f1a0661d506b0848c7563b317708bc42d48ff599ebc29e"],"state_sha256":"5e5e443c92b25ba3023c1e1e43a35ea5a66f1f4cc9fa91f2746a5a5711fce628"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"pAr7lUbBJSm8OP5kQa3103bA3xsqgS8fAzJEUj/HeaMP0frVqo9rIusRGyi9iK3xR3l7V60VLVlvRMBOpWZiCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T00:01:50.573395Z","bundle_sha256":"59f2d6998391ddd3a94b148eaecdbcbe2f4cccfb4c1f36914814e65942711470"}}