List your company
ResearchFormerly WizWand

Moral Reasoning on MMLU Moral Scenarios (test)

0.86AccuracySKIG
0.16320.34410.5250.7059May 21, 2024
Updated 8mo ago

Evaluation results

MethodLinks
SKIGLarge Language Model=GPT-42024.050.86
CoTLarge Language Model=GPT-42024.050.8
I/OLarge Language Model=GPT-42024.050.78
SKIGLarge Language Model=GPT-3.5 TURBO2024.050.71
TELarge Language Model=GPT-42024.050.6
SKIGLarge Language Model=MISTRAL-7B2024.050.58
TELarge Language Model=GPT-3.5 TURBO2024.050.54
CoTLarge Language Model=GPT-3.5 TURBO2024.050.52
SKIGLarge Language Model=DAVINCI2024.050.51
TELarge Language Model=MISTRAL-7B2024.050.5
I/OLarge Language Model=GPT-3.5 TURBO2024.050.42
CoTLarge Language Model=DAVINCI2024.050.39
I/OLarge Language Model=DAVINCI2024.050.38
I/OLarge Language Model=MISTRAL-7B2024.050.38
CoTLarge Language Model=MISTRAL-7B2024.050.37
TELarge Language Model=DAVINCI2024.050.35
TELarge Language Model=CURIE2024.050.28
SKIGLarge Language Model=BABBAGE2024.050.27
SKIGLarge Language Model=CURIE2024.050.26
I/OLarge Language Model=BABBAGE2024.050.25
SKIGLarge Language Model=ADA2024.050.24
I/OLarge Language Model=ADA2024.050.23
CoTLarge Language Model=ADA2024.050.23
CoTLarge Language Model=BABBAGE2024.050.21
CoTLarge Language Model=CURIE2024.050.21
TELarge Language Model=ADA2024.050.21
TELarge Language Model=BABBAGE2024.050.2
I/OLarge Language Model=CURIE2024.050.19