{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:3LK7G6NXHYQJTCR5NQ7BYNOBQX","short_pith_number":"pith:3LK7G6NX","schema_version":"1.0","canonical_sha256":"dad5f379b73e20998a3d6c3e1c35c185c88b78f0abc7293646dc1f65ecb02e5d","source":{"kind":"arxiv","id":"2004.14543","version":3},"attestation_state":"computed","paper":{"title":"TAVAT: Token-Aware Virtual Adversarial Training for Language Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Linyang Li, Xipeng Qiu","submitted_at":"2020-04-30T02:03:24Z","abstract_excerpt":"Gradient-based adversarial training is widely used in improving the robustness of neural networks, while it cannot be easily adapted to natural language processing tasks since the embedding space is discrete. In natural language processing fields, virtual adversarial training is introduced since texts are discrete and cannot be perturbed by gradients directly. Alternatively, virtual adversarial training, which generates perturbations on the embedding space, is introduced in NLP tasks. Despite its success, existing virtual adversarial training methods generate perturbations roughly constrained "},"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":"2004.14543","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-30T02:03:24Z","cross_cats_sorted":[],"title_canon_sha256":"6048efe1903e447395a7c725308155751a4acbe9e4a4f1185dbc9d08e3ccb202","abstract_canon_sha256":"8d7dad4954f71ef5782c3e27149c79d43335c03dfe02b29f49e9bec1335a5de9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:57:05.723440Z","signature_b64":"rLyfC7hNzAmfaqFvnGK/kKFSq487A3iWCxt4vSnlF7fRg6rBo9Yp3VUAIeBJyop8mN2lHglzwclZGYzmbCE1Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dad5f379b73e20998a3d6c3e1c35c185c88b78f0abc7293646dc1f65ecb02e5d","last_reissued_at":"2026-07-05T01:57:05.722930Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:57:05.722930Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TAVAT: Token-Aware Virtual Adversarial Training for Language Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Linyang Li, Xipeng Qiu","submitted_at":"2020-04-30T02:03:24Z","abstract_excerpt":"Gradient-based adversarial training is widely used in improving the robustness of neural networks, while it cannot be easily adapted to natural language processing tasks since the embedding space is discrete. In natural language processing fields, virtual adversarial training is introduced since texts are discrete and cannot be perturbed by gradients directly. Alternatively, virtual adversarial training, which generates perturbations on the embedding space, is introduced in NLP tasks. Despite its success, existing virtual adversarial training methods generate perturbations roughly constrained "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.14543","kind":"arxiv","version":3},"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/2004.14543/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":"2004.14543","created_at":"2026-07-05T01:57:05.722990+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.14543v3","created_at":"2026-07-05T01:57:05.722990+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.14543","created_at":"2026-07-05T01:57:05.722990+00:00"},{"alias_kind":"pith_short_12","alias_value":"3LK7G6NXHYQJ","created_at":"2026-07-05T01:57:05.722990+00:00"},{"alias_kind":"pith_short_16","alias_value":"3LK7G6NXHYQJTCR5","created_at":"2026-07-05T01:57:05.722990+00:00"},{"alias_kind":"pith_short_8","alias_value":"3LK7G6NX","created_at":"2026-07-05T01:57:05.722990+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.13439","citing_title":"S-GBT: Smooth Growth Bound Tensor for Certified Robustness Against Word Substitution Attacks in NLP","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX","json":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX.json","graph_json":"https://pith.science/api/pith-number/3LK7G6NXHYQJTCR5NQ7BYNOBQX/graph.json","events_json":"https://pith.science/api/pith-number/3LK7G6NXHYQJTCR5NQ7BYNOBQX/events.json","paper":"https://pith.science/paper/3LK7G6NX"},"agent_actions":{"view_html":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX","download_json":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX.json","view_paper":"https://pith.science/paper/3LK7G6NX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.14543&json=true","fetch_graph":"https://pith.science/api/pith-number/3LK7G6NXHYQJTCR5NQ7BYNOBQX/graph.json","fetch_events":"https://pith.science/api/pith-number/3LK7G6NXHYQJTCR5NQ7BYNOBQX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX/action/storage_attestation","attest_author":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX/action/author_attestation","sign_citation":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX/action/citation_signature","submit_replication":"https://pith.science/pith/3LK7G6NXHYQJTCR5NQ7BYNOBQX/action/replication_record"}},"created_at":"2026-07-05T01:57:05.722990+00:00","updated_at":"2026-07-05T01:57:05.722990+00:00"}