{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:D6N4SZT5FDJK7T7V3EIMZDKBFE","short_pith_number":"pith:D6N4SZT5","schema_version":"1.0","canonical_sha256":"1f9bc9667d28d2afcff5d910cc8d4129197bd37017747aa48371b711a7b2896b","source":{"kind":"arxiv","id":"2303.14342","version":2},"attestation_state":"computed","paper":{"title":"Analyzing the Performance of GPT-3.5 and GPT-4 in Grammatical Error Correction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Diana Galvan-Sosa, Keisuke Sakaguchi, Kentaro Inui, Michael Zock, Steven Coyne","submitted_at":"2023-03-25T03:08:49Z","abstract_excerpt":"GPT-3 and GPT-4 models are powerful, achieving high performance on a variety of Natural Language Processing tasks. However, there is a relative lack of detailed published analysis of their performance on the task of grammatical error correction (GEC). To address this, we perform experiments testing the capabilities of a GPT-3.5 model (text-davinci-003) and a GPT-4 model (gpt-4-0314) on major GEC benchmarks. We compare the performance of different prompts in both zero-shot and few-shot settings, analyzing intriguing or problematic outputs encountered with different prompt formats. We report the"},"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":"2303.14342","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-03-25T03:08:49Z","cross_cats_sorted":[],"title_canon_sha256":"396f1e4cdb27cfafa00252c0ee97da32f9c0389b10ccd2416f29708c71f58b81","abstract_canon_sha256":"d59d03f1d3ab7fbf33fb54466a8485f7fb89e2fb0749a7af5129388bebb1bff4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:15:27.529003Z","signature_b64":"WW950r5s5KUw7kdWSfxviD1/xlGoJrI4MZuIFqMI/4ei+HH5AhLKAMjUA/4oVxYTgANhhIM/93U2O430MomkAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1f9bc9667d28d2afcff5d910cc8d4129197bd37017747aa48371b711a7b2896b","last_reissued_at":"2026-07-05T06:15:27.528562Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:15:27.528562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analyzing the Performance of GPT-3.5 and GPT-4 in Grammatical Error Correction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Diana Galvan-Sosa, Keisuke Sakaguchi, Kentaro Inui, Michael Zock, Steven Coyne","submitted_at":"2023-03-25T03:08:49Z","abstract_excerpt":"GPT-3 and GPT-4 models are powerful, achieving high performance on a variety of Natural Language Processing tasks. However, there is a relative lack of detailed published analysis of their performance on the task of grammatical error correction (GEC). To address this, we perform experiments testing the capabilities of a GPT-3.5 model (text-davinci-003) and a GPT-4 model (gpt-4-0314) on major GEC benchmarks. We compare the performance of different prompts in both zero-shot and few-shot settings, analyzing intriguing or problematic outputs encountered with different prompt formats. We report the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.14342","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/2303.14342/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":"2303.14342","created_at":"2026-07-05T06:15:27.528620+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.14342v2","created_at":"2026-07-05T06:15:27.528620+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.14342","created_at":"2026-07-05T06:15:27.528620+00:00"},{"alias_kind":"pith_short_12","alias_value":"D6N4SZT5FDJK","created_at":"2026-07-05T06:15:27.528620+00:00"},{"alias_kind":"pith_short_16","alias_value":"D6N4SZT5FDJK7T7V","created_at":"2026-07-05T06:15:27.528620+00:00"},{"alias_kind":"pith_short_8","alias_value":"D6N4SZT5","created_at":"2026-07-05T06:15:27.528620+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.09378","citing_title":"APIO: Automatic Prompt Induction and Optimization for Grammatical Error Correction and Text Simplification","ref_index":2023,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE","json":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE.json","graph_json":"https://pith.science/api/pith-number/D6N4SZT5FDJK7T7V3EIMZDKBFE/graph.json","events_json":"https://pith.science/api/pith-number/D6N4SZT5FDJK7T7V3EIMZDKBFE/events.json","paper":"https://pith.science/paper/D6N4SZT5"},"agent_actions":{"view_html":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE","download_json":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE.json","view_paper":"https://pith.science/paper/D6N4SZT5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.14342&json=true","fetch_graph":"https://pith.science/api/pith-number/D6N4SZT5FDJK7T7V3EIMZDKBFE/graph.json","fetch_events":"https://pith.science/api/pith-number/D6N4SZT5FDJK7T7V3EIMZDKBFE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE/action/storage_attestation","attest_author":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE/action/author_attestation","sign_citation":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE/action/citation_signature","submit_replication":"https://pith.science/pith/D6N4SZT5FDJK7T7V3EIMZDKBFE/action/replication_record"}},"created_at":"2026-07-05T06:15:27.528620+00:00","updated_at":"2026-07-05T06:15:27.528620+00:00"}