List your company
ResearchFormerly WizWand

Uncertainty Quantification on MulFactTrap (test)

0.898ROC AUCRUfact
0.4863680.5932340.70010.806966Jan 1, 2026
Updated 8mo ago

Evaluation results

MethodLinks
RUfactModel=Mistral-7B2026.010.8980.74790.6684
RUgenModel=Mistral-7B2026.010.88760.78680.6584
RUgenModel=Vicuna-13B2026.010.8660.73140.5925
MaxprobmeanModel=ChatGLM3-6B-32K2026.010.85580.38620.4617
RUfactModel=Vicuna-13B2026.010.84790.69980.5214
RUfactModel=LlaMA3-8B-Instruct2026.010.81950.61560.5267
MaxprobmaxModel=ChatGLM3-6B-32K2026.010.81910.36580.4141
CCPmeanModel=ChatGLM3-6B-32K2026.010.80280.36220.393
CCPmaxModel=Mistral-7B2026.010.78110.40470.4775
RUgenModel=ChatGLM3-6B-32K2026.010.77750.46090.3644
RUfactModel=ChatGLM3-6B-32K2026.010.76470.42560.3518
CCPmaxModel=ChatGLM3-6B-32K2026.010.76080.28720.3384
MaxprobmaxModel=LlaMA3-8B-Instruct2026.010.75210.39570.4312
PEModel=Mistral-7B2026.010.74690.40250.4194
LN - PEModel=Mistral-7B2026.010.74670.40170.4191
CCPmeanModel=LlaMA3-8B-Instruct2026.010.74570.41960.4203
RUgenModel=LlaMA3-8B-Instruct2026.010.74420.54320.4075
MaxprobmeanModel=LlaMA3-8B-Instruct2026.010.7350.42890.4021
CCPmeanModel=Mistral-7B2026.010.73220.38790.3945
CCPmaxModel=LlaMA3-8B-Instruct2026.010.73010.36450.3936
MaxprobmeanModel=Mistral-7B2026.010.72490.34510.3822
MaxprobmaxModel=Mistral-7B2026.010.71060.29070.3577
MaxprobmeanModel=Vicuna-13B2026.010.70990.34590.3634
CCPmeanModel=Vicuna-13B2026.010.68350.34650.3178
PEModel=Vicuna-13B2026.010.67980.27370.2911
LN - PEModel=ChatGLM3-6B-32K2026.010.6770.20420.2324
MaxprobmaxModel=Vicuna-13B2026.010.66060.22510.2781
PEModel=ChatGLM3-6B-32K2026.010.65110.14070.1984
CCPmaxModel=Vicuna-13B2026.010.64910.24180.2582
PEModel=LlaMA3-8B-Instruct2026.010.53530.10530.0589
LN - PEModel=Vicuna-13B2026.010.5268-0.0956-0.0433
LN - PEModel=LlaMA3-8B-Instruct2026.010.50220.0561-0.0036