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

NCTE

Benchmarks

Task NameDataset NameSOTA ResultTrend
Rubric-based scoringNCTE 1-7 scoring scale (transcript segments)
MAE0.96
26
Automated Speech RecognitionNCTE (test)
WER16.54
6
Dialogue GenerationNCTE
BLEU8.99
5
Correlation of uptake scores with instructional outcomesNCTE
z-scored coefficient (beta)0.121
4
Conditional Hypothesis GenerationNCTE
Helpfulness3
2
Dialogue Quality EvaluationNCTE
BF1 (qt, at)34
1
Showing 6 of 6 rows