{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:NPAZVVVTHAOX6VBRIWTJOQ7SX5","short_pith_number":"pith:NPAZVVVT","schema_version":"1.0","canonical_sha256":"6bc19ad6b3381d7f543145a69743f2bf7a422fe904cea1c2f4841d7aa9d98211","source":{"kind":"arxiv","id":"2009.11264","version":2},"attestation_state":"computed","paper":{"title":"On the Ability and Limitations of Transformers to Recognize Formal Languages","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Kabir Ahuja, Navin Goyal, Satwik Bhattamishra","submitted_at":"2020-09-23T17:21:33Z","abstract_excerpt":"Transformers have supplanted recurrent models in a large number of NLP tasks. However, the differences in their abilities to model different syntactic properties remain largely unknown. Past works suggest that LSTMs generalize very well on regular languages and have close connections with counter languages. In this work, we systematically study the ability of Transformers to model such languages as well as the role of its individual components in doing so. We first provide a construction of Transformers for a subclass of counter languages, including well-studied languages such as n-ary Boolean"},"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":"2009.11264","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-23T17:21:33Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a08c86fb1fd3387f934744416840e9fd1a67924c1ad059e327f992a59ef61d2c","abstract_canon_sha256":"c3d548ea8980b579ffb4b04fe598768ca9f5501dd11fc775ef08af87f65bf251"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:41:20.244586Z","signature_b64":"D25ySLC6B12cjAiwa6nH/Cn6Jgt9sIPN7v3qLJpJ3KevMoky7KYZFLt2Pxu1lIdNTr2L0lwZHoxqRpkGPmtwDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6bc19ad6b3381d7f543145a69743f2bf7a422fe904cea1c2f4841d7aa9d98211","last_reissued_at":"2026-07-05T01:41:20.244080Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:41:20.244080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Ability and Limitations of Transformers to Recognize Formal Languages","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Kabir Ahuja, Navin Goyal, Satwik Bhattamishra","submitted_at":"2020-09-23T17:21:33Z","abstract_excerpt":"Transformers have supplanted recurrent models in a large number of NLP tasks. However, the differences in their abilities to model different syntactic properties remain largely unknown. Past works suggest that LSTMs generalize very well on regular languages and have close connections with counter languages. In this work, we systematically study the ability of Transformers to model such languages as well as the role of its individual components in doing so. We first provide a construction of Transformers for a subclass of counter languages, including well-studied languages such as n-ary Boolean"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2009.11264","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/2009.11264/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":"2009.11264","created_at":"2026-07-05T01:41:20.244134+00:00"},{"alias_kind":"arxiv_version","alias_value":"2009.11264v2","created_at":"2026-07-05T01:41:20.244134+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2009.11264","created_at":"2026-07-05T01:41:20.244134+00:00"},{"alias_kind":"pith_short_12","alias_value":"NPAZVVVTHAOX","created_at":"2026-07-05T01:41:20.244134+00:00"},{"alias_kind":"pith_short_16","alias_value":"NPAZVVVTHAOX6VBR","created_at":"2026-07-05T01:41:20.244134+00:00"},{"alias_kind":"pith_short_8","alias_value":"NPAZVVVT","created_at":"2026-07-05T01:41:20.244134+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00605","citing_title":"Looped Transformers with Layer Normalization Provably Learn the Power Method","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2507.12549","citing_title":"The Serial Scaling Hypothesis","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2602.01651","citing_title":"On the Spatiotemporal Dynamics of Generalization in Neural Networks","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2602.14814","citing_title":"Learning State-Tracking from Code Using Linear RNNs","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2603.29069","citing_title":"On the Mirage of Long-Range Dependency, with an Application to Integer Multiplication","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17857","citing_title":"On the Emergence of Syntax by Means of Local Interaction","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5","json":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5.json","graph_json":"https://pith.science/api/pith-number/NPAZVVVTHAOX6VBRIWTJOQ7SX5/graph.json","events_json":"https://pith.science/api/pith-number/NPAZVVVTHAOX6VBRIWTJOQ7SX5/events.json","paper":"https://pith.science/paper/NPAZVVVT"},"agent_actions":{"view_html":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5","download_json":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5.json","view_paper":"https://pith.science/paper/NPAZVVVT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2009.11264&json=true","fetch_graph":"https://pith.science/api/pith-number/NPAZVVVTHAOX6VBRIWTJOQ7SX5/graph.json","fetch_events":"https://pith.science/api/pith-number/NPAZVVVTHAOX6VBRIWTJOQ7SX5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5/action/storage_attestation","attest_author":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5/action/author_attestation","sign_citation":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5/action/citation_signature","submit_replication":"https://pith.science/pith/NPAZVVVTHAOX6VBRIWTJOQ7SX5/action/replication_record"}},"created_at":"2026-07-05T01:41:20.244134+00:00","updated_at":"2026-07-05T01:41:20.244134+00:00"}