Share your thoughts, 1 month free Claude Pro on usSee more

Grammatical Error Correction on BEA 2019 (test)

75.9F0.5

T5-XXL

Updated 1mo ago

Evaluation Results

Method
T5-XXL 2024.06	75.9	-	-
BART Baseline + CDA w/ denoising 2024.06	75.4	77.7	67.5
MultiTaskBART 2024.06	75.3	78.2	65.5
EPO (Llama2-7b-chat) 2026.05	74.6	78.8	61.6
TemplateGEC 2024.06	74.1	76.8	64.8
BART Baseline + CDA w/o denoising 2024.06	74.1	76.1	67.3
T5-XL 2024.06	73.9	-	-
GECToR (deberta-v3-large) 2026.05	73.9	79.3	58
BART Baseline 2024.06	73.5	74.5	68.9
T5 (t5-v1_1-large) 2026.05	73.4	76.9	62.3
EPO (Llama2-7b-chat) 2026.05	73.3	75.8	64.9
MixEdit 2024.06	73.2	76.4	62.7
ShallowAD 2024.06	72.9	-	-
SynGEC 2024.06	72.9	75.1	65.5
GECToR 2024.06	72.4	79.2	53.9
T5-large 2024.06	72.1	-	-
GECToR (bert-base-cased) 2026.05	70	77.3	50.9
EPO (Llama2-7b-chat) 2026.05	68.9	70.6	61.7
Qwen3-8B 2026.05	68.7	73.8	53.7
Qwen3-8B 2026.05	66.2	69.3	56.3
gemma-2-9b-it 2026.05	64	65	60.3
Llama-3.1-8B-Instruct 2026.05	63.1	66.9	51.5
gemma-2-9b-it 2026.05	59.5	58.4	64.4
gemma-2-9b-it 2026.05	59.2	58.3	62.8
Qwen3-8B 2026.05	56.3	54.8	62.9
Llama-3.1-8B-Instruct 2026.05	54	52.3	61.6
Llama-3.1-8B-Instruct 2026.05	52.1	50.5	59.3
Qwen3-8B + COCOGEC 2026.06	38.22	36.63	46.12
Qwen3-235B 2026.06	38.2	37.01	43.82
GPT-4o 2026.06	36.6	34.43	48.89
T5-large + COCOGEC 2026.06	32.64	32.93	31.52
GECToR-large + COCOGEC 2026.06	30.07	51.55	11.27
T5-large + TypeDA 2026.06	29.9	27.96	41.42
T5-large + DISCO 2026.06	29.43	27.7	39.28
Qwen3-8B + CPR method 2026.06	29.38	27.48	40.64
T5-large + CPR method 2026.06	28.99	27.29	38.7
T5-large 2026.06	28.92	27.27	38.22
Qwen3-14B 2026.06	28.7	26.86	39.57
GECToR-large + DISCO 2026.06	28.65	31.49	21.05
Qwen3-8B + TypeDA 2026.06	28.59	25.94	48.4
Qwen3-8B 2026.06	28.29	26.68	37.34
GECToR-large + CPR method 2026.06	28.16	28.32	27.55
Qwen3-8B + DISCO 2026.06	27.93	26.2	38.01
GECToR-large + TypeDA 2026.06	26.63	26.87	25.64
GECToR-large 2026.06	26.36	26.86	24.54
LLaMA3-8B 2026.06	25.55	23.09	44.52
Qwen3-4B 2026.06	22.13	22.14	22.11