{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AT4SW2FMZHPBZQO5HGVECIGRXO","short_pith_number":"pith:AT4SW2FM","schema_version":"1.0","canonical_sha256":"04f92b68acc9de1cc1dd39aa4120d1bba01b4a9706ed74a1e44d854c11d6158d","source":{"kind":"arxiv","id":"2505.11746","version":1},"attestation_state":"computed","paper":{"title":"Token Masking Improves Transformer-Based Text Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"John Bowen, Rojin Taheri, Xianglong Xu","submitted_at":"2025-05-16T23:06:11Z","abstract_excerpt":"While transformer-based models achieve strong performance on text classification, we explore whether masking input tokens can further enhance their effectiveness. We propose token masking regularization, a simple yet theoretically motivated method that randomly replaces input tokens with a special [MASK] token at probability p. This introduces stochastic perturbations during training, leading to implicit gradient averaging that encourages the model to capture deeper inter-token dependencies. Experiments on language identification and sentiment analysis -- across diverse models (mBERT, Qwen2.5-"},"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":"2505.11746","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-16T23:06:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"570cc34f6ec81f42e460d2127ef42463b050e488ea6a7588d337fb4d5af2c751","abstract_canon_sha256":"3542195c2dfc4c798a7281f20233b2f7396e8d8e11f8bfc8c959dd84e8cfa3c6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:04:26.041020Z","signature_b64":"NME2LW+A2wO1/3MobP5RJG5jysX994yaCaJ/lCGKQISL10qgOmFPSalbsRm18FEWMJN5Aiqh7WGQ0lPnvtKpAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"04f92b68acc9de1cc1dd39aa4120d1bba01b4a9706ed74a1e44d854c11d6158d","last_reissued_at":"2026-07-05T11:04:26.040500Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:04:26.040500Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Token Masking Improves Transformer-Based Text Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"John Bowen, Rojin Taheri, Xianglong Xu","submitted_at":"2025-05-16T23:06:11Z","abstract_excerpt":"While transformer-based models achieve strong performance on text classification, we explore whether masking input tokens can further enhance their effectiveness. We propose token masking regularization, a simple yet theoretically motivated method that randomly replaces input tokens with a special [MASK] token at probability p. This introduces stochastic perturbations during training, leading to implicit gradient averaging that encourages the model to capture deeper inter-token dependencies. Experiments on language identification and sentiment analysis -- across diverse models (mBERT, Qwen2.5-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11746","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/2505.11746/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":"2505.11746","created_at":"2026-07-05T11:04:26.040582+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.11746v1","created_at":"2026-07-05T11:04:26.040582+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11746","created_at":"2026-07-05T11:04:26.040582+00:00"},{"alias_kind":"pith_short_12","alias_value":"AT4SW2FMZHPB","created_at":"2026-07-05T11:04:26.040582+00:00"},{"alias_kind":"pith_short_16","alias_value":"AT4SW2FMZHPBZQO5","created_at":"2026-07-05T11:04:26.040582+00:00"},{"alias_kind":"pith_short_8","alias_value":"AT4SW2FM","created_at":"2026-07-05T11:04:26.040582+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/AT4SW2FMZHPBZQO5HGVECIGRXO","json":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO.json","graph_json":"https://pith.science/api/pith-number/AT4SW2FMZHPBZQO5HGVECIGRXO/graph.json","events_json":"https://pith.science/api/pith-number/AT4SW2FMZHPBZQO5HGVECIGRXO/events.json","paper":"https://pith.science/paper/AT4SW2FM"},"agent_actions":{"view_html":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO","download_json":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO.json","view_paper":"https://pith.science/paper/AT4SW2FM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.11746&json=true","fetch_graph":"https://pith.science/api/pith-number/AT4SW2FMZHPBZQO5HGVECIGRXO/graph.json","fetch_events":"https://pith.science/api/pith-number/AT4SW2FMZHPBZQO5HGVECIGRXO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO/action/storage_attestation","attest_author":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO/action/author_attestation","sign_citation":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO/action/citation_signature","submit_replication":"https://pith.science/pith/AT4SW2FMZHPBZQO5HGVECIGRXO/action/replication_record"}},"created_at":"2026-07-05T11:04:26.040582+00:00","updated_at":"2026-07-05T11:04:26.040582+00:00"}