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

Language Model Calibration on SCITLDR (dev)

0.072smECE

ACUTE cosine Platt

Updated 1mo ago

Evaluation Results

Method	Links
ACUTE cosine Platt 2026.06		0.072	0.678	0.721	0.925	0.774
ACUTE pca20 Isotonic 2026.06		0.072	0.692	0.724	0.925	0.78
ACUTE pca20 Beta 2026.06		0.072	0.699	0.723	0.925	0.782
ACUTE cosine Beta 2026.06		0.073	0.678	0.721	0.925	0.774
ACUTE pca20 Platt 2026.06		0.073	0.698	0.723	0.925	0.782
ACUTE pca10 Platt 2026.06		0.074	0.694	0.722	0.925	0.779
ACUTE pca10 Beta 2026.06		0.074	0.694	0.722	0.925	0.779
ACUTE pca10 Isotonic 2026.06		0.075	0.686	0.723	0.925	0.777
ACUTE cosine Isotonic 2026.06		0.078	0.673	0.718	0.924	0.771
Platt 2026.06		0.079	0.668	0.721	0.925	0.771
Isotonic 2026.06		0.079	0.669	0.721	0.925	0.771
Beta 2026.06		0.079	0.668	0.721	0.925	0.771
HRE 2026.06		0.081	0.665	0.721	0.925	0.769
NWKR 2026.06		0.082	0.667	0.721	0.925	0.77
ACUTE mid act 2026.06		0.084	0.689	0.722	0.925	0.778
ACUTE late act 2026.06		0.084	0.681	0.722	0.925	0.775
ACUTE cosine 2026.06		0.085	0.669	0.717	0.925	0.769
ACUTE early act 2026.06		0.086	0.686	0.722	0.925	0.777
ACUTE pca10 2026.06		0.089	0.685	0.723	0.925	0.777
ACUTE pca20 2026.06		0.091	0.688	0.723	0.925	0.778
Raw Conf 2026.06		0.152	0.366	0.721	0.925	0.667