| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| PKU-SafeRLHF prompts n = 100 samples (Sev-Low) | IPO Entropic τ=10 | CVaR(0.125) Combined Win Rate53.6 | 10 | 21d ago | |
| PKU-SafeRLHF Sev-3 prompts n = 100 samples | IPO Entropic τ=10 | Combined Win Rate (CVaR 0.125)60.9 | 10 | 21d ago | |
| PKU-SafeRLHF Conflict prompts n = 100 samples | IPO Entropic τ=10 | CVaR(0.125) Combined Win Rate40.3 | 10 | 21d ago | |
| PKU-SafeRLHF Random prompts n = 100 samples | IPO Entropic τ=10 | CVaR(0.125) Combined Win Rate37.1 | 10 | 21d ago |