Unsupervised GEC + cLang8

Reported on 4 benchmarks across 1 task

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing4 results

Grammatical Error CorrectiononCoNLL-2014 Shared Task
F0.5
69.6
best: 72.8 (Ensembles of best 7 models + GRECO + GTP-rerank)
Grammatical Error CorrectiononCoNLL-2014 Shared Task
Precision
75
best: 83.9 (Ensembles of best 7 models + GRECO + GTP-rerank)
Grammatical Error CorrectiononCoNLL-2014 Shared Task
Recall
53.8
Grammatical Error CorrectiononBEA-2019 (test)
F0.5
76.5
best: 81.4 (Majority-voting ensemble on best 7 models)