List your company
ResearchFormerly WizWand

Schwartz Value (3-class) Evaluation on ValEval (Original)

76.29Accuracy (3-class)Mistral-7b
14.066830.220946.37562.5291Jul 15, 2024
Updated 8mo ago

Evaluation results

MethodLinks
Mistral-7bEvaluation Protocol=Tuning-based Evaluation, Backbone=Mistral-7b2024.0776.29
CLAVE-MistralEvaluation Protocol=Tuning-based Evaluation, Backbone=Mistral-7b2024.0775.26
Phi-3Evaluation Protocol=Tuning-based Evaluation, Backbone=Phi-32024.0771.93
CLAVE-LlamaEvaluation Protocol=Tuning-based Evaluation, Backbone=Llama2-7b2024.0769.85
GPT-2-LargeEvaluation Protocol=Tuning-based Evaluation, Backbone=GPT-2-Large2024.0769.02
Llama2-7bEvaluation Protocol=Tuning-based Evaluation, Backbone=Llama2-7b2024.0764.26
CrowdworkerEvaluation Protocol=Tuning-based Evaluation2024.0760.21
CalibrateEvaluation Protocol=Prompt-based Evaluation2024.0755.53
VanillaEvaluation Protocol=Prompt-based Evaluation2024.0754.79
Few-shotEvaluation Protocol=Prompt-based Evaluation2024.0754.29
Chain-of-thoughtEvaluation Protocol=Prompt-based Evaluation2024.0754.04
AllureEvaluation Protocol=Prompt-based Evaluation2024.0753.59
G-EvalEvaluation Protocol=Prompt-based Evaluation2024.0752.76
FairEvalEvaluation Protocol=Prompt-based Evaluation2024.0740.83
WideDeepEvaluation Protocol=Prompt-based Evaluation2024.0725
ChatEvalEvaluation Protocol=Prompt-based Evaluation2024.0716.46