List your company

Social harm risk evaluation on SHARP (n=901 prompts)

1.689CVaR95 Log-Riskclaude-sonnet-4.5
1.420683.231845.0436.85416Jan 29, 2026

Evaluation results

MethodLinks
claude-sonnet-4.52026.011.6890.1580.772
gemini-1.5-pro2026.013.4990.4630.962
claude-3.5-sonnet2026.013.7820.4490.965
gemini-2.5-pro2026.013.8680.560.971
qwen3-235bParameters=235b2026.013.8950.5890.967
gpt-4o2026.014.5190.7480.984
gpt-oss-120bParameters=120b2026.015.2861.2090.994
deepseek-chat2026.015.7351.2960.994
mistral-large2026.015.9881.3220.995
llama3-3-70bParameters=70b2026.017.9341.9680.998
llama3-1-405bParameters=405b2026.018.3972.1540.998