{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:4TTKOMB6EVPVYNIW563PPOSH2D","short_pith_number":"pith:4TTKOMB6","schema_version":"1.0","canonical_sha256":"e4e6a7303e255f5c3516efb6f7ba47d0fec556fa1426107154a368619f8d99f6","source":{"kind":"arxiv","id":"2505.00926","version":3},"attestation_state":"computed","paper":{"title":"How Transformers Learn Regular Language Recognition: A Theoretical Study on Training Dynamics and Implicit Bias","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jing Yang, Ruiquan Huang, Yingbin Liang","submitted_at":"2025-05-02T00:07:35Z","abstract_excerpt":"Language recognition tasks are fundamental in natural language processing (NLP) and have been widely used to benchmark the performance of large language models (LLMs). These tasks also play a crucial role in explaining the working mechanisms of transformers. In this work, we focus on two representative tasks in the category of regular language recognition, known as `even pairs' and `parity check', the aim of which is to determine whether the occurrences of certain subsequences in a given sequence are even. Our goal is to explore how a one-layer transformer, consisting of an attention layer fol"},"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":"2505.00926","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-02T00:07:35Z","cross_cats_sorted":["cs.CL","stat.ML"],"title_canon_sha256":"6d64e73ded2287e293c5e0b4583dec4da8991a1fc783a99ea8ee56427a66d836","abstract_canon_sha256":"cfe964c145d40bd9a98b265332f283f72030112f643b9274478074d0df03246f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:44.465939Z","signature_b64":"o04UVmbJ8kinmH4AXLZgqNt0FtSxocjBaqMbJd3Ek5rsN4XXqgK539gJEFCCZ6BtZbQ9cOgOKda8bCyocW2fBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4e6a7303e255f5c3516efb6f7ba47d0fec556fa1426107154a368619f8d99f6","last_reissued_at":"2026-07-05T11:11:44.465458Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:44.465458Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Transformers Learn Regular Language Recognition: A Theoretical Study on Training Dynamics and Implicit Bias","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jing Yang, Ruiquan Huang, Yingbin Liang","submitted_at":"2025-05-02T00:07:35Z","abstract_excerpt":"Language recognition tasks are fundamental in natural language processing (NLP) and have been widely used to benchmark the performance of large language models (LLMs). These tasks also play a crucial role in explaining the working mechanisms of transformers. In this work, we focus on two representative tasks in the category of regular language recognition, known as `even pairs' and `parity check', the aim of which is to determine whether the occurrences of certain subsequences in a given sequence are even. Our goal is to explore how a one-layer transformer, consisting of an attention layer fol"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.00926","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/2505.00926/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":"2505.00926","created_at":"2026-07-05T11:11:44.465512+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.00926v3","created_at":"2026-07-05T11:11:44.465512+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.00926","created_at":"2026-07-05T11:11:44.465512+00:00"},{"alias_kind":"pith_short_12","alias_value":"4TTKOMB6EVPV","created_at":"2026-07-05T11:11:44.465512+00:00"},{"alias_kind":"pith_short_16","alias_value":"4TTKOMB6EVPVYNIW","created_at":"2026-07-05T11:11:44.465512+00:00"},{"alias_kind":"pith_short_8","alias_value":"4TTKOMB6","created_at":"2026-07-05T11:11:44.465512+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00183","citing_title":"Agentic Transformers Provably Learn to Search via Reinforcement Learning","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D","json":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D.json","graph_json":"https://pith.science/api/pith-number/4TTKOMB6EVPVYNIW563PPOSH2D/graph.json","events_json":"https://pith.science/api/pith-number/4TTKOMB6EVPVYNIW563PPOSH2D/events.json","paper":"https://pith.science/paper/4TTKOMB6"},"agent_actions":{"view_html":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D","download_json":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D.json","view_paper":"https://pith.science/paper/4TTKOMB6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.00926&json=true","fetch_graph":"https://pith.science/api/pith-number/4TTKOMB6EVPVYNIW563PPOSH2D/graph.json","fetch_events":"https://pith.science/api/pith-number/4TTKOMB6EVPVYNIW563PPOSH2D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D/action/storage_attestation","attest_author":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D/action/author_attestation","sign_citation":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D/action/citation_signature","submit_replication":"https://pith.science/pith/4TTKOMB6EVPVYNIW563PPOSH2D/action/replication_record"}},"created_at":"2026-07-05T11:11:44.465512+00:00","updated_at":"2026-07-05T11:11:44.465512+00:00"}