{"id":"https://openalex.org/W7083468041","doi":"https://doi.org/10.48550/arxiv.2509.20811","title":"Leveraging What's Overfixed: Post-Correction via LLM Grammatical Error Overcorrection","display_name":"Leveraging What's Overfixed: Post-Correction via LLM Grammatical Error Overcorrection","publication_year":2025,"publication_date":"2025-09-25","ids":{"openalex":"https://openalex.org/W7083468041","doi":"https://doi.org/10.48550/arxiv.2509.20811"},"language":"en","primary_location":{"id":"doi:10.48550/arxiv.2509.20811","doi":"https://doi.org/10.48550/arxiv.2509.20811","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.20811","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2509.20811","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Park, Taehee","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Park, Taehee","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Do, Heejin","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Do, Heejin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Lee, Gary Geunbae","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Gary Geunbae","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9829000234603882,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.9829000234603882,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.0071000000461936,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.003700000001117587,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/grammatical-error-correction","display_name":"grammatical error correction","score":0.996999979019165},{"id":"https://openalex.org/keywords/overcorrection","display_name":"overcorrection","score":0.8840000033378601},{"id":"https://openalex.org/keywords/post-correction","display_name":"post-correction","score":0.8080000281333923},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.800000011920929},{"id":"https://openalex.org/keywords/small-language-models","display_name":"small language models","score":0.949999988079071},{"id":"https://openalex.org/keywords/fine-tuning","display_name":"fine-tuning","score":0.7179999947547913},{"id":"https://openalex.org/keywords/precision-recall-trade-off","display_name":"precision-recall trade-off","score":0.8650000095367432},{"id":"https://openalex.org/keywords/poco","display_name":"POCO","score":0.527999997138977}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6747000217437744},{"id":"https://openalex.org/C100660578","wikidata":"https://www.wikidata.org/wiki/Q18733","display_name":"Recall","level":2,"score":0.6567999720573425},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.6504999995231628},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.5238999724388123},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.478300005197525},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.4413999915122986},{"id":"https://openalex.org/C81669768","wikidata":"https://www.wikidata.org/wiki/Q2359161","display_name":"Precision and recall","level":2,"score":0.39969998598098755},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3982999920845032},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3779999911785126},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.35760000348091125},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.3490000069141388},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3343000113964081},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.3253999948501587},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.2759999930858612},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.2759999930858612},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2734000086784363}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2509.20811","doi":"https://doi.org/10.48550/arxiv.2509.20811","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.20811","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2509.20811","doi":"https://doi.org/10.48550/arxiv.2509.20811","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.20811","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[{"score":0.4940214157104492,"display_name":"Quality education","id":"https://metadata.un.org/sdg/4"}],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Robust":[0],"supervised":[1,120],"fine-tuned":[2],"small":[3],"Language":[4,27],"Models":[5,28],"(sLMs)":[6],"often":[7,30],"show":[8,31],"high":[9,17],"reliability":[10,117],"but":[11],"tend":[12],"to":[13,39,49,78,95,103],"undercorrect.":[14],"They":[15],"achieve":[16],"precision":[18],"at":[19],"the":[20,32,45,51,109,116,140],"cost":[21],"of":[22,47,112,118,143],"low":[23,40],"recall.":[24],"Conversely,":[25],"Large":[26],"(LLMs)":[29],"opposite":[33],"tendency,":[34],"making":[35],"excessive":[36],"overcorrection,":[37],"leading":[38],"precision.":[41,70],"To":[42],"effectively":[43,128],"harness":[44],"strengths":[46],"LLMs":[48,113],"address":[50],"recall":[52,68,80,134],"challenges":[53],"in":[54],"sLMs,":[55],"we":[56],"propose":[57],"Post-Correction":[58],"via":[59,76,91],"Overcorrection":[60],"(PoCO),":[61],"a":[62,87],"novel":[63],"approach":[64],"that":[65,126],"strategically":[66],"balances":[67,129],"and":[69,97],"PoCO":[71,127],"first":[72],"intentionally":[73],"triggers":[74],"overcorrection":[75],"LLM":[77],"maximize":[79],"by":[81,107,132],"allowing":[82],"comprehensive":[83],"revisions,":[84],"then":[85],"applies":[86],"targeted":[88],"post-correction":[89],"step":[90],"fine-tuning":[92],"smaller":[93,119],"models":[94],"identify":[96],"refine":[98],"erroneous":[99],"outputs.":[100],"We":[101],"aim":[102],"harmonize":[104],"both":[105],"aspects":[106],"leveraging":[108],"generative":[110],"power":[111],"while":[114],"preserving":[115],"models.":[121],"Our":[122],"extensive":[123],"experiments":[124],"demonstrate":[125],"GEC":[130],"performance":[131],"increasing":[133],"with":[135],"competitive":[136],"precision,":[137],"ultimately":[138],"improving":[139],"overall":[141],"quality":[142],"grammatical":[144],"error":[145],"correction.":[146]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2025-10-10T00:00:00"}