Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Helpfulness, Honesty, and Harmlessness Alignment Evaluation on BBH HHH

95Harmlessness Score

GPT-4

55.37665.66375.9586.237Aug 18, 2023
Updated 4d ago

Evaluation Results

MethodLinks
2023.08
9585809187
2023.08
9585809187
2023.08
89.665.559.383.774.5
2023.08
8175.481.38681
2023.08
8154.167.87970.1
2023.08
70.75964.474.467.1
2023.08
68.95955.969.763.4
2023.08
68.968.864.479.170.3
2023.08
68.962.355.972.164.8
2023.08
56.957.454.272.160.2