List your company
ResearchFormerly WizWand

Social Risks (2-class) Evaluation on ValEval (Original)

88.57Accuracy (Social Risks 2-class)Mistral-7b
79.251681.670884.0986.5092Jul 15, 2024
Updated 8mo ago

Evaluation results

MethodLinks
Mistral-7bEvaluation Protocol=Tuning-based Evaluation, Backbone=Mistral-7b2024.0788.57
CLAVE-MistralEvaluation Protocol=Tuning-based Evaluation, Backbone=Mistral-7b2024.0788.36
CrowdworkerEvaluation Protocol=Tuning-based Evaluation2024.0786
GPT-2-LargeEvaluation Protocol=Tuning-based Evaluation, Backbone=GPT-2-Large2024.0785.86
FairEvalEvaluation Protocol=Prompt-based Evaluation2024.0785.83
AllureEvaluation Protocol=Prompt-based Evaluation2024.0785.66
CalibrateEvaluation Protocol=Prompt-based Evaluation2024.0785.2
CLAVE-LlamaEvaluation Protocol=Tuning-based Evaluation, Backbone=Llama2-7b2024.0785.03
VanillaEvaluation Protocol=Prompt-based Evaluation2024.0784.89
Phi-3Evaluation Protocol=Tuning-based Evaluation, Backbone=Phi-32024.0784.83
ChatEvalEvaluation Protocol=Prompt-based Evaluation2024.0783.9
Llama2-7bEvaluation Protocol=Tuning-based Evaluation, Backbone=Llama2-7b2024.0783.57
Chain-of-thoughtEvaluation Protocol=Prompt-based Evaluation2024.0783.25
WideDeepEvaluation Protocol=Prompt-based Evaluation2024.0782.5
G-EvalEvaluation Protocol=Prompt-based Evaluation2024.0781.88
Few-shotEvaluation Protocol=Prompt-based Evaluation2024.0779.61