{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:B7RDMKHK6OR56J3ITYYM2QX5A6","short_pith_number":"pith:B7RDMKHK","canonical_record":{"source":{"id":"2211.04898","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-09T14:03:22Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8153be6c094a63a26671379d9f5aa87342555c18a565635ab2fc0098a75b0717","abstract_canon_sha256":"943e4b1a8651546ae19f3fa4641ab417bb3e37b4a0fa672941c0afad5ee0887a"},"schema_version":"1.0"},"canonical_sha256":"0fe23628eaf3a3df27689e30cd42fd07884f3e15c318d0fe0668e6291a959ecb","source":{"kind":"arxiv","id":"2211.04898","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2211.04898","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"arxiv_version","alias_value":"2211.04898v2","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.04898","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"pith_short_12","alias_value":"B7RDMKHK6OR5","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"pith_short_16","alias_value":"B7RDMKHK6OR56J3I","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"pith_short_8","alias_value":"B7RDMKHK","created_at":"2026-07-05T05:16:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:B7RDMKHK6OR56J3ITYYM2QX5A6","target":"record","payload":{"canonical_record":{"source":{"id":"2211.04898","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-09T14:03:22Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8153be6c094a63a26671379d9f5aa87342555c18a565635ab2fc0098a75b0717","abstract_canon_sha256":"943e4b1a8651546ae19f3fa4641ab417bb3e37b4a0fa672941c0afad5ee0887a"},"schema_version":"1.0"},"canonical_sha256":"0fe23628eaf3a3df27689e30cd42fd07884f3e15c318d0fe0668e6291a959ecb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:16:27.351912Z","signature_b64":"lvImafcNns+/d0D6bsfU+YXBMHm+jHLvXUuNpNB1U3J19RuLfzM9vb5wGbh2TDClUjpvOa/WZIvabheVa2etAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0fe23628eaf3a3df27689e30cd42fd07884f3e15c318d0fe0668e6291a959ecb","last_reissued_at":"2026-07-05T05:16:27.351426Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:16:27.351426Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2211.04898","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:16:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"boEMRD4FrXCnsXU41HYmxqC3dLdpiUZSKgZ0GDDikvpPQuYswGG3K6fN5dB4F4p0mSQU3JWdXCicEMbDJqU4Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T10:21:26.705625Z"},"content_sha256":"d48213cd92cbfc4aa99717f731f9eec9afecd7c5203404d46ed710d5889ef081","schema_version":"1.0","event_id":"sha256:d48213cd92cbfc4aa99717f731f9eec9afecd7c5203404d46ed710d5889ef081"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:B7RDMKHK6OR56J3ITYYM2QX5A6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Mask More and Mask Later: Efficient Pre-training of Masked Language Models by Disentangling the [MASK] Token","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Baohao Liao, Christof Monz, David Thulke, Hermann Ney, Sanjika Hewavitharana","submitted_at":"2022-11-09T14:03:22Z","abstract_excerpt":"The pre-training of masked language models (MLMs) consumes massive computation to achieve good results on downstream NLP tasks, resulting in a large carbon footprint. In the vanilla MLM, the virtual tokens, [MASK]s, act as placeholders and gather the contextualized information from unmasked tokens to restore the corrupted information. It raises the question of whether we can append [MASK]s at a later layer, to reduce the sequence length for earlier layers and make the pre-training more efficient. We show: (1) [MASK]s can indeed be appended at a later layer, being disentangled from the word emb"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.04898","kind":"arxiv","version":2},"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/2211.04898/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":"2026-07-05T05:16:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cn1NFHTcr7nx8JEhmlWi3NLevN+Tva+RTEc22BFXvNjNXw731yOZm5OU8z/+ao7cA5NY7LMYDj4T3+9iPSYMBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T10:21:26.706575Z"},"content_sha256":"48a4f5e12702b53d8815f8aaffb99e253602e9c6c6085c86ce6271d05c92e2b6","schema_version":"1.0","event_id":"sha256:48a4f5e12702b53d8815f8aaffb99e253602e9c6c6085c86ce6271d05c92e2b6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/B7RDMKHK6OR56J3ITYYM2QX5A6/bundle.json","state_url":"https://pith.science/pith/B7RDMKHK6OR56J3ITYYM2QX5A6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/B7RDMKHK6OR56J3ITYYM2QX5A6/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":"2026-08-11T10:21:26Z","links":{"resolver":"https://pith.science/pith/B7RDMKHK6OR56J3ITYYM2QX5A6","bundle":"https://pith.science/pith/B7RDMKHK6OR56J3ITYYM2QX5A6/bundle.json","state":"https://pith.science/pith/B7RDMKHK6OR56J3ITYYM2QX5A6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/B7RDMKHK6OR56J3ITYYM2QX5A6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:B7RDMKHK6OR56J3ITYYM2QX5A6","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":"943e4b1a8651546ae19f3fa4641ab417bb3e37b4a0fa672941c0afad5ee0887a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-09T14:03:22Z","title_canon_sha256":"8153be6c094a63a26671379d9f5aa87342555c18a565635ab2fc0098a75b0717"},"schema_version":"1.0","source":{"id":"2211.04898","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2211.04898","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"arxiv_version","alias_value":"2211.04898v2","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.04898","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"pith_short_12","alias_value":"B7RDMKHK6OR5","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"pith_short_16","alias_value":"B7RDMKHK6OR56J3I","created_at":"2026-07-05T05:16:27Z"},{"alias_kind":"pith_short_8","alias_value":"B7RDMKHK","created_at":"2026-07-05T05:16:27Z"}],"graph_snapshots":[{"event_id":"sha256:48a4f5e12702b53d8815f8aaffb99e253602e9c6c6085c86ce6271d05c92e2b6","target":"graph","created_at":"2026-07-05T05:16:27Z","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/2211.04898/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The pre-training of masked language models (MLMs) consumes massive computation to achieve good results on downstream NLP tasks, resulting in a large carbon footprint. In the vanilla MLM, the virtual tokens, [MASK]s, act as placeholders and gather the contextualized information from unmasked tokens to restore the corrupted information. It raises the question of whether we can append [MASK]s at a later layer, to reduce the sequence length for earlier layers and make the pre-training more efficient. We show: (1) [MASK]s can indeed be appended at a later layer, being disentangled from the word emb","authors_text":"Baohao Liao, Christof Monz, David Thulke, Hermann Ney, Sanjika Hewavitharana","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-09T14:03:22Z","title":"Mask More and Mask Later: Efficient Pre-training of Masked Language Models by Disentangling the [MASK] Token"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.04898","kind":"arxiv","version":2},"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:d48213cd92cbfc4aa99717f731f9eec9afecd7c5203404d46ed710d5889ef081","target":"record","created_at":"2026-07-05T05:16:27Z","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":"943e4b1a8651546ae19f3fa4641ab417bb3e37b4a0fa672941c0afad5ee0887a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-11-09T14:03:22Z","title_canon_sha256":"8153be6c094a63a26671379d9f5aa87342555c18a565635ab2fc0098a75b0717"},"schema_version":"1.0","source":{"id":"2211.04898","kind":"arxiv","version":2}},"canonical_sha256":"0fe23628eaf3a3df27689e30cd42fd07884f3e15c318d0fe0668e6291a959ecb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0fe23628eaf3a3df27689e30cd42fd07884f3e15c318d0fe0668e6291a959ecb","first_computed_at":"2026-07-05T05:16:27.351426Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:16:27.351426Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lvImafcNns+/d0D6bsfU+YXBMHm+jHLvXUuNpNB1U3J19RuLfzM9vb5wGbh2TDClUjpvOa/WZIvabheVa2etAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T05:16:27.351912Z","signed_message":"canonical_sha256_bytes"},"source_id":"2211.04898","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d48213cd92cbfc4aa99717f731f9eec9afecd7c5203404d46ed710d5889ef081","sha256:48a4f5e12702b53d8815f8aaffb99e253602e9c6c6085c86ce6271d05c92e2b6"],"state_sha256":"7ec817e4338aab20427281b0e79a37509a8a76b4932cbd2017d1299c2bc81d45"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sZh8APe1duRvw7WI4H6vCYSiOFII4Zop8Y1oZqKWk+xjJAfbTWL/4Urqu0brb0Yog1klu6r5gUmCFajuccvNCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-11T10:21:26.712707Z","bundle_sha256":"e625a3ba414d22563dd4fa2bb425ca06a4e6eaaecaa1c12fc5c532ca18013254"}}