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

CodeWorkout

Benchmarks

Task NameDataset NameSOTA ResultTrend
Knowledge EstimationCodeWorkout
Accuracy72.32
49
Code SimilarityCodeWorkout (5-fold averaged)
CodeBLEU0.593
49
Computerized Adaptive TestingCodeWorkout (test)
Exp.% (med)41.4
7
Knowledge TracingCodeWorkout (test)
A1 Performance Score82.4
4
Predictive Modeling (AFM)CodeWorkout Selected KC set (test)
AUC63.1
3
Predictive Modeling (AFM)CodeWorkout Generated KC set (test)
AUC0.676
3
Predictive Modeling (AFM)CodeWorkout Human KC set (test)
AUC66.5
3
Learning Curve fitCodeWorkout Selected KC set (test)
RMSE0.069
3
Learning Curve fitCodeWorkout Generated KC set (test)
RMSE0.07
3
Learning Curve fitCodeWorkout Human KC set (test)
RMSE0.077
3
Showing 10 of 10 rows