List your company
ResearchFormerly WizWand

Predicting Language Model Output on PH

0.0073Brier ScoreLAMP
-0.0133160.1258420.2650.404158May 17, 2025
Updated 6mo ago

Evaluation results

MethodLinks
LAMPBase Language Model=Gemini 2.5 Flash2025.050.0073
LAMPBase Language Model=GPT-4.1-mini2025.050.0193
LAMPBase Language Model=Mistral Large2025.050.0209
SHAPBase Language Model=GPT-4.1-mini2025.050.0383
LAMPBase Language Model=Claude 3.5 Haiku2025.050.0759
Uniform baselineBase Language Model=Claude 3.5 Haiku2025.050.1167
SHAPBase Language Model=Mistral Large2025.050.1258
ANCHORSBase Language Model=Mistral Large2025.050.1406
Uniform baselineBase Language Model=GPT-4.1-mini2025.050.1421
Uniform baselineBase Language Model=Mistral Large2025.050.1471
ANCHORSBase Language Model=GPT-4.1-mini2025.050.148
SHAPBase Language Model=Claude 3.5 Haiku2025.050.1559
ANCHORSBase Language Model=Claude 3.5 Haiku2025.050.1702
ANCHORSBase Language Model=Gemini 2.5 Flash2025.050.1841
Mean baselineBase Language Model=GPT-4.1-mini2025.050.2025
Mean baselineBase Language Model=Claude 3.5 Haiku2025.050.2246
LIMEBase Language Model=Gemini 2.5 Flash2025.050.2417
SHAPBase Language Model=Gemini 2.5 Flash2025.050.2455
LIMEBase Language Model=Claude 3.5 Haiku2025.050.2625
Mean baselineBase Language Model=Mistral Large2025.050.2848
Mean baselineBase Language Model=Gemini 2.5 Flash2025.050.3646
Uniform baselineBase Language Model=Gemini 2.5 Flash2025.050.3646
LIMEBase Language Model=Mistral Large2025.050.3944
Random baselineBase Language Model=Claude 3.5 Haiku2025.050.414
Random baselineBase Language Model=GPT-4.1-mini2025.050.4194
LIMEBase Language Model=GPT-4.1-mini2025.050.4241
Random baselineBase Language Model=Mistral Large2025.050.427
Random baselineBase Language Model=Gemini 2.5 Flash2025.050.5227