{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:ID3ZDPQREJSQTB2ZAXKC4KGSYJ","short_pith_number":"pith:ID3ZDPQR","schema_version":"1.0","canonical_sha256":"40f791be11226509875905d42e28d2c250dc40b5572a8ff75770e5fe040a1b0e","source":{"kind":"arxiv","id":"1808.09031","version":1},"attestation_state":"computed","paper":{"title":"Targeted Syntactic Evaluation of Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Rebecca Marvin, Tal Linzen","submitted_at":"2018-08-27T20:42:51Z","abstract_excerpt":"We present a dataset for evaluating the grammaticality of the predictions of a language model. We automatically construct a large number of minimally different pairs of English sentences, each consisting of a grammatical and an ungrammatical sentence. The sentence pairs represent different variations of structure-sensitive phenomena: subject-verb agreement, reflexive anaphora and negative polarity items. We expect a language model to assign a higher probability to the grammatical sentence than the ungrammatical one. In an experiment using this data set, an LSTM language model performed poorly "},"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":"1808.09031","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-27T20:42:51Z","cross_cats_sorted":[],"title_canon_sha256":"0dac79b3de1506467510f8719b7f4c229d54222a2b9eadc3619144775c4cfa75","abstract_canon_sha256":"e7e850783e6dc3558509d5d8e1847196c3a97daeaea711c31eeb7b4422020020"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:07:04.807472Z","signature_b64":"IP7mgcAf9Xked1BxICklt/vIuVvtGuXjBkcwUWoCOHrnPAfVWIwPRkZJr2fetafoL27A7Qq/yscEw6vcLkbCAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"40f791be11226509875905d42e28d2c250dc40b5572a8ff75770e5fe040a1b0e","last_reissued_at":"2026-05-18T00:07:04.806793Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:07:04.806793Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Targeted Syntactic Evaluation of Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Rebecca Marvin, Tal Linzen","submitted_at":"2018-08-27T20:42:51Z","abstract_excerpt":"We present a dataset for evaluating the grammaticality of the predictions of a language model. We automatically construct a large number of minimally different pairs of English sentences, each consisting of a grammatical and an ungrammatical sentence. The sentence pairs represent different variations of structure-sensitive phenomena: subject-verb agreement, reflexive anaphora and negative polarity items. We expect a language model to assign a higher probability to the grammatical sentence than the ungrammatical one. In an experiment using this data set, an LSTM language model performed poorly "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.09031","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":""},"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":"1808.09031","created_at":"2026-05-18T00:07:04.806891+00:00"},{"alias_kind":"arxiv_version","alias_value":"1808.09031v1","created_at":"2026-05-18T00:07:04.806891+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.09031","created_at":"2026-05-18T00:07:04.806891+00:00"},{"alias_kind":"pith_short_12","alias_value":"ID3ZDPQREJSQ","created_at":"2026-05-18T12:32:28.185984+00:00"},{"alias_kind":"pith_short_16","alias_value":"ID3ZDPQREJSQTB2Z","created_at":"2026-05-18T12:32:28.185984+00:00"},{"alias_kind":"pith_short_8","alias_value":"ID3ZDPQR","created_at":"2026-05-18T12:32:28.185984+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/ID3ZDPQREJSQTB2ZAXKC4KGSYJ","json":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ.json","graph_json":"https://pith.science/api/pith-number/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/graph.json","events_json":"https://pith.science/api/pith-number/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/events.json","paper":"https://pith.science/paper/ID3ZDPQR"},"agent_actions":{"view_html":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ","download_json":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ.json","view_paper":"https://pith.science/paper/ID3ZDPQR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1808.09031&json=true","fetch_graph":"https://pith.science/api/pith-number/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/graph.json","fetch_events":"https://pith.science/api/pith-number/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/action/storage_attestation","attest_author":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/action/author_attestation","sign_citation":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/action/citation_signature","submit_replication":"https://pith.science/pith/ID3ZDPQREJSQTB2ZAXKC4KGSYJ/action/replication_record"}},"created_at":"2026-05-18T00:07:04.806891+00:00","updated_at":"2026-05-18T00:07:04.806891+00:00"}