List your company
ResearchFormerly WizWand

Schwartz Value (3-class) Evaluation on ValEval Generalized

88.91AccuracyCrowdworker
66.102872.023977.94583.8661Jul 15, 2024
Updated 8mo ago

Evaluation results

MethodLinks
CrowdworkerEvaluation Protocol=Tuning-based Evaluation2024.0788.91
CLAVE-LlamaEvaluation Protocol=Tuning-based Evaluation, Backbone=Llama2-7b2024.0783.71
CLAVE-MistralEvaluation Protocol=Tuning-based Evaluation, Backbone=Mistral-7b2024.0782.45
FairEvalEvaluation Protocol=Prompt-based Evaluation2024.0782.35
ChatEvalEvaluation Protocol=Prompt-based Evaluation2024.0782.35
WideDeepEvaluation Protocol=Prompt-based Evaluation2024.0782.35
Mistral-7bEvaluation Protocol=Tuning-based Evaluation, Backbone=Mistral-7b2024.0778.19
Llama2-7bEvaluation Protocol=Tuning-based Evaluation, Backbone=Llama2-7b2024.0777.69
GPT-2-LargeEvaluation Protocol=Tuning-based Evaluation, Backbone=GPT-2-Large2024.0777.36
Chain-of-thoughtEvaluation Protocol=Prompt-based Evaluation2024.0773.68
Phi-3Evaluation Protocol=Tuning-based Evaluation, Backbone=Phi-32024.0772.93
VanillaEvaluation Protocol=Prompt-based Evaluation2024.0771.62
CalibrateEvaluation Protocol=Prompt-based Evaluation2024.0770.74
G-EvalEvaluation Protocol=Prompt-based Evaluation2024.0769.36
AllureEvaluation Protocol=Prompt-based Evaluation2024.0767.86
Few-shotEvaluation Protocol=Prompt-based Evaluation2024.0766.98