{"url":"/task/grammatical-error-correction","name":"Grammatical Error Correction","slug":"grammatical-error-correction","description_markdown":"Grammatical Error Correction (GEC) is the task of correcting different kinds of errors in text such as spelling, punctuation, grammatical, and word choice errors. \r\n\r\nGEC is typically formulated as a sentence correction task. A GEC system takes a potentially erroneous sentence as input and is expected to transform it to its corrected version. See the example given below: \r\n\r\n| Input (Erroneous)          | Output (Corrected)     |\r\n| -------------------------  | ---------------------- |\r\n|She see Tom is catched by policeman in park at last night. | She saw Tom caught by a policeman in the park last night.|","categories":[{"name":"Natural Language Processing","url":"/area/natural-language-processing"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"archive_url"},"counts":{"papers_tagged":415,"papers_with_code":142,"benchmarks":13,"benchmark_tables_in_archive":13,"benchmark_tables_shown":13,"benchmark_tables_withheld_as_spam":0,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":16,"subtasks":1,"parent_tasks":0},"benchmarks":[{"leaderboard":"/sota/grammatical-error-correction-on-conll-2014","slug":"grammatical-error-correction-on-conll-2014","dataset":"CoNLL-2014 Shared Task","dataset_url":"/dataset/conll-1","rows_in_archive":23,"metrics":["F0.5","Precision","Recall"],"first_row_in_archive_order":{"model":"Ensembles of best 7 models + GRECO + GTP-rerank","paper_title":"Pillars of Grammatical Error Correction: Comprehensive Inspection Of Contemporary Approaches In The Era of Large Language Models","paper_url":"/paper/pillars-of-grammatical-error-correction","paper_date":"2024-04-23","arxiv_id":"2404.14914","code_links":[{"title":"grammarly/pillars-of-gec","url":"https://github.com/grammarly/pillars-of-gec"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-bea-2019-test","slug":"grammatical-error-correction-on-bea-2019-test","dataset":"BEA-2019 (test)","dataset_url":"/dataset/locness-corpus","rows_in_archive":19,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"Majority-voting ensemble on best 7 models","paper_title":"Pillars of Grammatical Error Correction: Comprehensive Inspection Of Contemporary Approaches In The Era of Large Language Models","paper_url":"/paper/pillars-of-grammatical-error-correction","paper_date":"2024-04-23","arxiv_id":"2404.14914","code_links":[{"title":"grammarly/pillars-of-gec","url":"https://github.com/grammarly/pillars-of-gec"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-falko-merlin","slug":"grammatical-error-correction-on-falko-merlin","dataset":"Falko-MERLIN","dataset_url":null,"rows_in_archive":6,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"Llama + 1M BT + gold","paper_title":"To Err Is Human, but Llamas Can Learn It Too","paper_url":"/paper/to-err-is-human-but-llamas-can-learn-it-too","paper_date":"2024-03-08","arxiv_id":"2403.05493","code_links":[{"title":"TartuNLP/gec-llm","url":"https://github.com/TartuNLP/gec-llm"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-jfleg","slug":"grammatical-error-correction-on-jfleg","dataset":"JFLEG","dataset_url":"/dataset/jfleg","rows_in_archive":6,"metrics":["GLEU"],"first_row_in_archive_order":{"model":"VERNet","paper_title":"Neural Quality Estimation with Multiple Hypotheses for Grammatical Error Correction","paper_url":"/paper/neural-quality-estimation-with-multiple","paper_date":"2021-05-10","arxiv_id":"2105.04443","code_links":[{"title":"thunlp/VERNet","url":"https://github.com/thunlp/VERNet"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-ua-gec","slug":"grammatical-error-correction-on-ua-gec","dataset":"UA-GEC","dataset_url":"/dataset/ua-gec","rows_in_archive":5,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"Llama + 1M BT + gold","paper_title":"To Err Is Human, but Llamas Can Learn It Too","paper_url":"/paper/to-err-is-human-but-llamas-can-learn-it-too","paper_date":"2024-03-08","arxiv_id":"2403.05493","code_links":[{"title":"TartuNLP/gec-llm","url":"https://github.com/TartuNLP/gec-llm"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-restricted","slug":"grammatical-error-correction-on-restricted","dataset":"Restricted","dataset_url":"/dataset/jfleg","rows_in_archive":4,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"CNN Seq2Seq","paper_title":"A Multilayer Convolutional Encoder-Decoder Neural Network for Grammatical Error Correction","paper_url":"/paper/a-multilayer-convolutional-encoder-decoder","paper_date":"2018-01-26","arxiv_id":"1801.08831","code_links":[{"title":"nusnlp/mlconvgec2018","url":"https://github.com/nusnlp/mlconvgec2018"},{"title":"seaweiqing/neuraltalk_plus_charcnn","url":"https://github.com/seaweiqing/neuraltalk_plus_charcnn"},{"title":"seaweiqing/image2story","url":"https://github.com/seaweiqing/image2story"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-conll-2014-1","slug":"grammatical-error-correction-on-conll-2014-1","dataset":"CoNLL-2014 Shared Task (10 annotations)","dataset_url":"/dataset/conll-1","rows_in_archive":3,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"GRECO (vote+ESC)","paper_title":"System Combination via Quality Estimation for Grammatical Error Correction","paper_url":"/paper/system-combination-via-quality-estimation-for","paper_date":"2023-10-23","arxiv_id":"2310.14947","code_links":[{"title":"nusnlp/greco","url":"https://github.com/nusnlp/greco"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-unrestricted","slug":"grammatical-error-correction-on-unrestricted","dataset":"Unrestricted","dataset_url":"/dataset/jfleg","rows_in_archive":3,"metrics":["F0.5","GLEU"],"first_row_in_archive_order":{"model":"CNN Seq2Seq + Fluency Boost","paper_title":"Reaching Human-level Performance in Automatic Grammatical Error Correction: An Empirical Study","paper_url":"/paper/reaching-human-level-performance-in-automatic","paper_date":"2018-07-03","arxiv_id":"1807.01270","code_links":[{"title":"getao/human-performance-gec","url":"https://github.com/getao/human-performance-gec"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-_restricted_","slug":"grammatical-error-correction-on-_restricted_","dataset":"_Restricted_","dataset_url":"/dataset/jfleg","rows_in_archive":2,"metrics":["GLEU"],"first_row_in_archive_order":{"model":"Transformer","paper_title":"Approaching Neural Grammatical Error Correction as a Low-Resource Machine Translation Task","paper_url":"/paper/approaching-neural-grammatical-error","paper_date":"2018-04-16","arxiv_id":"1804.05940","code_links":[{"title":"grammatical/neural-naacl2018","url":"https://github.com/grammatical/neural-naacl2018"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-estgec-l2","slug":"grammatical-error-correction-on-estgec-l2","dataset":"EstGEC-L2","dataset_url":null,"rows_in_archive":1,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"Llama + 1M BT + gold","paper_title":"To Err Is Human, but Llamas Can Learn It Too","paper_url":"/paper/to-err-is-human-but-llamas-can-learn-it-too","paper_date":"2024-03-08","arxiv_id":"2403.05493","code_links":[{"title":"TartuNLP/gec-llm","url":"https://github.com/TartuNLP/gec-llm"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-fcgec","slug":"grammatical-error-correction-on-fcgec","dataset":"FCGEC","dataset_url":"/dataset/fcgec","rows_in_archive":1,"metrics":["exact match","F0.5"],"first_row_in_archive_order":{"model":"STG-Joint","paper_title":"FCGEC: Fine-Grained Corpus for Chinese Grammatical Error Correction","paper_url":"/paper/fcgec-fine-grained-corpus-for-chinese","paper_date":"2022-10-22","arxiv_id":"2210.12364","code_links":[{"title":"xlxwalex/FCGEC","url":"https://github.com/xlxwalex/FCGEC"},{"title":"xlxwalex/hycxg","url":"https://github.com/xlxwalex/hycxg"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-mucgec","slug":"grammatical-error-correction-on-mucgec","dataset":"MuCGEC","dataset_url":"/dataset/mucgec","rows_in_archive":1,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"GEC-DI (LM+GED)","paper_title":"Improving Seq2Seq Grammatical Error Correction via Decoding Interventions","paper_url":"/paper/improving-seq2seq-grammatical-error","paper_date":"2023-10-23","arxiv_id":"2310.14534","code_links":[{"title":"Jacob-Zhou/gecdi","url":"https://github.com/Jacob-Zhou/gecdi"}],"syntology":null}},{"leaderboard":"/sota/grammatical-error-correction-on-wi-locness-1","slug":"grammatical-error-correction-on-wi-locness-1","dataset":"WI-LOCNESS","dataset_url":"/dataset/locness-corpus","rows_in_archive":1,"metrics":["F0.5"],"first_row_in_archive_order":{"model":"RedPenNet","paper_title":"RedPenNet for Grammatical Error Correction: Outputs to Tokens, Attentions to Spans","paper_url":"/paper/redpennet-for-grammatical-error-correction","paper_date":"2023-09-19","arxiv_id":"2309.10898","code_links":[{"title":"webspellchecker/unlp-2023-shared-task","url":"https://github.com/webspellchecker/unlp-2023-shared-task"}],"syntology":null}}],"datasets":[{"url":"/dataset/conll-1","name":"CoNLL","full_name":"","num_papers_in_archive":187},{"url":"/dataset/jfleg","name":"JFLEG","full_name":"JHU FLuency-Extended GUG corpus","num_papers_in_archive":91},{"url":"/dataset/conll-2014-shared-task-grammatical-error","name":"CoNLL-2014 Shared Task: Grammatical Error Correction","full_name":"jamal","num_papers_in_archive":79},{"url":"/dataset/locness-corpus","name":"WI-LOCNESS","full_name":"Cambridge English Write & Improve & LOCNESS","num_papers_in_archive":29},{"url":"/dataset/mucgec","name":"MuCGEC","full_name":"Multi-Reference Multi-Source Evaluation Dataset for Chinese Grammatical Error Correction","num_papers_in_archive":15},{"url":"/dataset/akces-gec","name":"AKCES-GEC","full_name":"","num_papers_in_archive":11},{"url":"/dataset/ua-gec","name":"UA-GEC","full_name":"UA-GEC: Grammatical Error Correction and Fluency Corpus for the Ukrainian Language","num_papers_in_archive":9},{"url":"/dataset/gmeg-wiki","name":"GMEG-wiki","full_name":"","num_papers_in_archive":4},{"url":"/dataset/github-typo-corpus","name":"GitHub Typo Corpus","full_name":"GitHub Typo Corpus","num_papers_in_archive":3},{"url":"/dataset/gmeg-yahoo","name":"GMEG-yahoo","full_name":"","num_papers_in_archive":2},{"url":"/dataset/fcgec","name":"FCGEC","full_name":"FCGEC: Fine-Grained Corpus for Chinese Grammatical Error Correction","num_papers_in_archive":1},{"url":"/dataset/kor-lang8","name":"Kor-Lang8","full_name":"Lang-8 Korean Corpus","num_papers_in_archive":1},{"url":"/dataset/kor-learner","name":"Kor-Learner","full_name":"Korean Learner Corpus","num_papers_in_archive":1},{"url":"/dataset/kor-native","name":"Kor-Native","full_name":"Native Korean Corpus","num_papers_in_archive":1},{"url":"/dataset/nasgec","name":"NaSGEC","full_name":"","num_papers_in_archive":1},{"url":"/dataset/the-write-improve-corpus-2024","name":"The Write & Improve Corpus 2024","full_name":"","num_papers_in_archive":1}],"subtasks":[{"url":"/task/grammatical-error-detection","name":"Grammatical Error Detection"}],"parent_tasks":[],"papers":{"order":"repositories listed in the archive (desc), then date (desc); the archive holds no stars","population":"papers tagged with this task that list at least one repository in the archive","shown":30,"of":142,"tagged_in_all":415,"items":[{"url":"/paper/improving-grammatical-error-correction-via","title":"Improving Grammatical Error Correction via Pre-Training a Copy-Augmented Architecture with Unlabeled Data","date":"2019-03-01","arxiv_id":"1903.00138","repositories_listed":6,"syntology":null},{"url":"/paper/lm-critic-language-models-for-unsupervised","title":"LM-Critic: Language Models for Unsupervised Grammatical Error Correction","date":"2021-09-14","arxiv_id":"2109.06822","repositories_listed":3,"syntology":null},{"url":"/paper/gector-grammatical-error-correction-tag-not","title":"GECToR -- Grammatical Error Correction: Tag, Not Rewrite","date":"2020-05-26","arxiv_id":"2005.12592","repositories_listed":3,"syntology":{"n":12,"n_ran":2,"n_unverified":10,"n_pointer_only":0}},{"url":"/paper/a-multilayer-convolutional-encoder-decoder","title":"A Multilayer Convolutional Encoder-Decoder Neural Network for Grammatical Error Correction","date":"2018-01-26","arxiv_id":"1801.08831","repositories_listed":3,"syntology":null},{"url":"/paper/multi-head-sequence-tagging-model-for","title":"Multi-head Sequence Tagging Model for Grammatical Error Correction","date":"2024-10-21","arxiv_id":"2410.16473","repositories_listed":2,"syntology":null},{"url":"/paper/revisiting-meta-evaluation-for-grammatical","title":"Revisiting Meta-evaluation for Grammatical Error Correction","date":"2024-03-05","arxiv_id":"2403.02674","repositories_listed":2,"syntology":null},{"url":"/paper/an-extended-sequence-tagging-vocabulary-for","title":"An Extended Sequence Tagging Vocabulary for Grammatical Error Correction","date":"2023-02-12","arxiv_id":"2302.05913","repositories_listed":2,"syntology":null},{"url":"/paper/fcgec-fine-grained-corpus-for-chinese","title":"FCGEC: Fine-Grained Corpus for Chinese Grammatical Error Correction","date":"2022-10-22","arxiv_id":"2210.12364","repositories_listed":2,"syntology":null},{"url":"/paper/linguistic-rules-based-corpus-generation-for","title":"Linguistic Rules-Based Corpus Generation for Native Chinese Grammatical Error Correction","date":"2022-10-19","arxiv_id":"2210.10442","repositories_listed":2,"syntology":null},{"url":"/paper/chinese-grammatical-error-correction-based-on-1","title":"Chinese grammatical error correction based on knowledge distillation","date":"2022-07-31","arxiv_id":"2208.00351","repositories_listed":2,"syntology":null},{"url":"/paper/mining-error-templates-for-grammatical-error","title":"Mining Error Templates for Grammatical Error Correction","date":"2022-06-23","arxiv_id":"2206.11569","repositories_listed":2,"syntology":null},{"url":"/paper/lossless-acceleration-for-seq2seq-generation","title":"Lossless Acceleration for Seq2seq Generation with Aggressive Decoding","date":"2022-05-20","arxiv_id":"2205.10350","repositories_listed":2,"syntology":null},{"url":"/paper/mucgec-a-multi-reference-multi-source-1","title":"MuCGEC: a Multi-Reference Multi-Source Evaluation Dataset for Chinese Grammatical Error Correction","date":"2022-04-23","arxiv_id":"2204.10994","repositories_listed":2,"syntology":null},{"url":"/paper/automatic-error-type-annotation-for-arabic","title":"Automatic Error Type Annotation for Arabic","date":"2021-09-16","arxiv_id":"2109.08068","repositories_listed":2,"syntology":null},{"url":"/paper/a-simple-recipe-for-multilingual-grammatical","title":"A Simple Recipe for Multilingual Grammatical Error Correction","date":"2021-06-07","arxiv_id":"2106.03830","repositories_listed":2,"syntology":null},{"url":"/paper/stronger-baselines-for-grammatical-error","title":"Stronger Baselines for Grammatical Error Correction Using Pretrained Encoder-Decoder Model","date":"2020-05-24","arxiv_id":"2005.11849","repositories_listed":2,"syntology":null},{"url":"/paper/a-neural-grammatical-error-correction-system","title":"A Neural Grammatical Error Correction System Built On Better Pre-training and Sequential Transfer Learning","date":"2019-07-02","arxiv_id":"1907.01256","repositories_listed":2,"syntology":null},{"url":"/paper/the-unreasonable-effectiveness-of-transformer","title":"The Unreasonable Effectiveness of Transformer Language Models in Grammatical Error Correction","date":"2019-06-04","arxiv_id":"1906.01733","repositories_listed":2,"syntology":null},{"url":"/paper/neural-network-translation-models-for","title":"Neural Network Translation Models for Grammatical Error Correction","date":"2016-06-01","arxiv_id":"1606.00189","repositories_listed":2,"syntology":null},{"url":"/paper/gec-metrics-a-unified-library-for-grammatical","title":"gec-metrics: A Unified Library for Grammatical Error Correction Evaluation","date":"2025-05-26","arxiv_id":"2505.19388","repositories_listed":1,"syntology":{"n":5,"n_ran":0,"n_unverified":5,"n_pointer_only":0}},{"url":"/paper/exploring-the-feasibility-of-multilingual","title":"Exploring the Feasibility of Multilingual Grammatical Error Correction with a Single LLM up to 9B parameters: A Comparative Study of 17 Models","date":"2025-05-09","arxiv_id":"2505.06004","repositories_listed":1,"syntology":null},{"url":"/paper/rethinking-evaluation-metrics-for-grammatical","title":"Rethinking Evaluation Metrics for Grammatical Error Correction: Why Use a Different Evaluation Process than Human?","date":"2025-02-13","arxiv_id":"2502.09416","repositories_listed":1,"syntology":{"n":3,"n_ran":0,"n_unverified":3,"n_pointer_only":0}},{"url":"/paper/explanation-based-in-context-demonstrations","title":"Explanation based In-Context Demonstrations Retrieval for Multilingual Grammatical Error Correction","date":"2025-02-12","arxiv_id":"2502.08507","repositories_listed":1,"syntology":{"n":5,"n_ran":0,"n_unverified":5,"n_pointer_only":0}},{"url":"/paper/dsgram-dynamic-weighting-sub-metrics-for","title":"DSGram: Dynamic Weighting Sub-Metrics for Grammatical Error Correction in the Era of Large Language Models","date":"2024-12-17","arxiv_id":"2412.12832","repositories_listed":1,"syntology":null},{"url":"/paper/improving-explainability-of-sentence-level","title":"Improving Explainability of Sentence-level Metrics via Edit-level Attribution for Grammatical Error Correction","date":"2024-12-17","arxiv_id":"2412.13110","repositories_listed":1,"syntology":null},{"url":"/paper/enhancing-grammatical-error-detection-using","title":"Enhancing Grammatical Error Detection using BERT with Cleaned Lang-8 Dataset","date":"2024-11-23","arxiv_id":"2411.15523","repositories_listed":1,"syntology":null},{"url":"/paper/efficient-and-interpretable-grammatical-error","title":"Efficient and Interpretable Grammatical Error Correction with Mixture of Experts","date":"2024-10-30","arxiv_id":"2410.23507","repositories_listed":1,"syntology":null},{"url":"/paper/a-simple-yet-effective-corpus-construction-1","title":"A Simple Yet Effective Corpus Construction Framework for Indonesian Grammatical Error Correction","date":"2024-10-28","arxiv_id":"2410.20838","repositories_listed":1,"syntology":null},{"url":"/paper/cleme2-0-towards-more-interpretable","title":"CLEME2.0: Towards More Interpretable Evaluation by Disentangling Edits for Grammatical Error Correction","date":"2024-07-01","arxiv_id":"2407.00934","repositories_listed":1,"syntology":null},{"url":"/paper/improving-grammatical-error-correction-via-1","title":"Improving Grammatical Error Correction via Contextual Data Augmentation","date":"2024-06-25","arxiv_id":"2406.17456","repositories_listed":1,"syntology":null}],"syntology_records":4,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; our fixture could not drive it","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"}},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"}}