BEA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
BEA shared task 2019 (test)
80.84F0.5 Score
139
BEA 2019 (test)
75.9F0.5
47
BEA 2019 (dev)
63.4F0.5 Score
19
BEA (dev)
71.7Precision (%)
14
BEA 19
74.7Precision
12
BEA (test)
87.5Precision
9
BEA (val)
48.72F0.5
8
BEA 2026 (test)
0.742RMSE (ES)
7
BEA Spanish Closed Track 2026
0.903RMSE
4
BEA Chinese Closed Track 2026
0.776RMSE
4
BEA German Closed Track 2026
0.885RMSE
4
BEA 2019 (dev)
72.13F0.5 Score
4
BEA (466/201 split)
0.309Standard Deviation (SD)
3
BEA Shared Task 2 Track 4 (Unseen Questions) 2026 (Evaluation)
0.49Quadratic Weighted Kappa
3
BEA (pooled items)
0.309Standard Deviation (SD)
2