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

Question Answering on TruthfulQA (Corr, Incorr, NΔ)

96.81Correlation Score

MAUVE

Updated 7d ago

Evaluation Results

Method	Links
MAUVE 2026.05		96.81	86.6	10.21
BERTScore 2026.05		83.8	81.29	2.51
EmbSim 2026.05		70.77	63.11	3.83
SimCSE 2026.05		68.98	55.77	6.61
MATCHA 2026.05		62.5	15.85	23.33
R1-F1 2026.05		45.12	38.87	6.25
METEOR 2026.05		42.08	33.6	8.48
RL-F1 2026.05		40.72	35.78	4.94
R2-F1 2026.05		24.4	21.53	2.87
BLEURT 2026.05		-29.73	-63.46	16.87