{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:OQDGX6TVEATBWV6XPPB7IMDV27","short_pith_number":"pith:OQDGX6TV","schema_version":"1.0","canonical_sha256":"74066bfa7520261b57d77bc3f43075d7efd5a1246550b8a29dc03ac125741bcf","source":{"kind":"arxiv","id":"2109.00442","version":1},"attestation_state":"computed","paper":{"title":"Position Masking for Improved Layout-Aware Document Understanding","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Anik Saha, Ashish Verma, Catherine Finegan-Dollak","submitted_at":"2021-09-01T15:40:15Z","abstract_excerpt":"Natural language processing for document scans and PDFs has the potential to enormously improve the efficiency of business processes. Layout-aware word embeddings such as LayoutLM have shown promise for classification of and information extraction from such documents. This paper proposes a new pre-training task called that can improve performance of layout-aware word embeddings that incorporate 2-D position embeddings. We compare models pre-trained with only language masking against models pre-trained with both language masking and position masking, and we find that position masking improves p"},"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":"2109.00442","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2021-09-01T15:40:15Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"6b6f6fc2d7428991c745aac9123721f3f7c1baa7436e602f730ca0b01917fb73","abstract_canon_sha256":"72e05486985a75e71cdfe6a6c2c7a81d67bd4754356104aaf92f66372e2e5e40"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:10:46.590900Z","signature_b64":"VKa78LjFuWZ2mr6G57oc3ii2/jB14thoYwKB/KXG2XgX0C2uVhnGweuJdxApwaOCfNVPk7D2+En9+4pYZjK5Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"74066bfa7520261b57d77bc3f43075d7efd5a1246550b8a29dc03ac125741bcf","last_reissued_at":"2026-07-05T03:10:46.590527Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:10:46.590527Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Position Masking for Improved Layout-Aware Document Understanding","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Anik Saha, Ashish Verma, Catherine Finegan-Dollak","submitted_at":"2021-09-01T15:40:15Z","abstract_excerpt":"Natural language processing for document scans and PDFs has the potential to enormously improve the efficiency of business processes. Layout-aware word embeddings such as LayoutLM have shown promise for classification of and information extraction from such documents. This paper proposes a new pre-training task called that can improve performance of layout-aware word embeddings that incorporate 2-D position embeddings. We compare models pre-trained with only language masking against models pre-trained with both language masking and position masking, and we find that position masking improves p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2109.00442","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/2109.00442/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":"2109.00442","created_at":"2026-07-05T03:10:46.590596+00:00"},{"alias_kind":"arxiv_version","alias_value":"2109.00442v1","created_at":"2026-07-05T03:10:46.590596+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2109.00442","created_at":"2026-07-05T03:10:46.590596+00:00"},{"alias_kind":"pith_short_12","alias_value":"OQDGX6TVEATB","created_at":"2026-07-05T03:10:46.590596+00:00"},{"alias_kind":"pith_short_16","alias_value":"OQDGX6TVEATBWV6X","created_at":"2026-07-05T03:10:46.590596+00:00"},{"alias_kind":"pith_short_8","alias_value":"OQDGX6TV","created_at":"2026-07-05T03:10:46.590596+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/OQDGX6TVEATBWV6XPPB7IMDV27","json":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27.json","graph_json":"https://pith.science/api/pith-number/OQDGX6TVEATBWV6XPPB7IMDV27/graph.json","events_json":"https://pith.science/api/pith-number/OQDGX6TVEATBWV6XPPB7IMDV27/events.json","paper":"https://pith.science/paper/OQDGX6TV"},"agent_actions":{"view_html":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27","download_json":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27.json","view_paper":"https://pith.science/paper/OQDGX6TV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2109.00442&json=true","fetch_graph":"https://pith.science/api/pith-number/OQDGX6TVEATBWV6XPPB7IMDV27/graph.json","fetch_events":"https://pith.science/api/pith-number/OQDGX6TVEATBWV6XPPB7IMDV27/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27/action/storage_attestation","attest_author":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27/action/author_attestation","sign_citation":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27/action/citation_signature","submit_replication":"https://pith.science/pith/OQDGX6TVEATBWV6XPPB7IMDV27/action/replication_record"}},"created_at":"2026-07-05T03:10:46.590596+00:00","updated_at":"2026-07-05T03:10:46.590596+00:00"}