List your company
ResearchFormerly WizWand

Safety Evaluation on HONEST

1.8ScoreUniversal Self-consistency
-0.0720.4140.91.386Jul 20, 2025
Updated 7mo ago

Evaluation results

MethodLinks
Universal Self-consistencyGroup=Baselines, Base Model=GPT-3.5-Turbo2025.071.8
Self-refineGroup=Baselines, Base Model=GPT-3.5-Turbo2025.071.3
Zero-shot-CoTGroup=Baselines, Base Model=GPT-3.5-Turbo2025.071.1
MADGroup=Baselines, Base Model=GPT-3.5-Turbo2025.070.9
ExpertPromptingGroup=Baselines, Base Model=GPT-3.5-Turbo2025.070.8
MEPGroup=Baselines, Base Model=GPT-3.5-Turbo2025.070.7
MetaCrit w/o ϕ↑Group=Ablation, Base Model=GPT-3.5-Turbo2025.070
MetaCrit w/o ϕ↓Group=Ablation, Base Model=GPT-3.5-Turbo2025.070
MetaCrit + GPT-3.5-TurboGroup=Ours, Base Model=GPT-3.5-Turbo2025.070
MetaCrit + DeepSeek-v3Group=Ours, Base Model=DeepSeek-v32025.070
MetaCrit + Claude-3.5-SonnetGroup=Ours, Base Model=Claude-3.5-Sonnet2025.070
MetaCrit + GPT-4oGroup=Ours, Base Model=GPT-4o2025.070