# Combined Win Rate Evaluation on PKU-SafeRLHF Sev-3 (n=100) Source: [Combined Win Rate Evaluation on PKU-SafeRLHF Sev-3 (n=100)]() Collection index: [llms.txt]() Data updated: 2026-05-12T14:02:24.616Z Evaluation of model performance using the Combined Win Rate metric, specifically calculated via CVaR(0.125), applied to the Sev-3 subset (n=100 samples) of the PKU-SafeRLHF dataset. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Combined win rate evaluation | | Dataset (including any stated split) | PKU-SafeRLHF Sev-3 prompts n = 100 samples | | Metric | Combined Win Rate (CVaR 0.125) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 10 of 10 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Combined Win Rate (CVaR 0.125)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Combined Win Rate (CVaR 0.125) | Result date | Report paper | | --- | --- | --- | --- | | IPO Entropic τ=10 | 60.9 | Not recorded | 6a02b491dbc836b87d824d17 | | IPO k = 1 Neutral | 57.2 | Not recorded | 6a02b491dbc836b87d824d17 | | EG Entropic τ=5 | 57.2 | Not recorded | 6a02b491dbc836b87d824d17 | | IPO CVaR α=0.125 | 56.8 | Not recorded | 6a02b491dbc836b87d824d17 | | IPO CVaR α=0.25 | 56.5 | Not recorded | 6a02b491dbc836b87d824d17 | | EG k = 1 Neutral | 54.9 | Not recorded | 6a02b491dbc836b87d824d17 | | IPO Entropic τ=5 | 54.8 | Not recorded | 6a02b491dbc836b87d824d17 | | NMD k = 1 Neutral | 52.5 | Not recorded | 6a02b491dbc836b87d824d17 | | EGPO† (HF) | 11 | Not recorded | 6a02b491dbc836b87d824d17 | | SFT Base | 1.9 | Not recorded | 6a02b491dbc836b87d824d17 | ### IPO Entropic τ=10 Combined Win Rate (CVaR 0.125): 60.9. Result ID: 6a03328b47f34ce924e7db12. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"tau (τ)":"10","objective":"Entropic"}. This record does not assert independent reproduction. ### IPO k = 1 Neutral Combined Win Rate (CVaR 0.125): 57.2. Result ID: 6a03328b47f34ce924e7db0e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"k":"1","objective":"Neutral"}. This record does not assert independent reproduction. ### EG Entropic τ=5 Combined Win Rate (CVaR 0.125): 57.2. Result ID: 6a03328b47f34ce924e7db13. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"tau (τ)":"5","objective":"Entropic"}. This record does not assert independent reproduction. ### IPO CVaR α=0.125 Combined Win Rate (CVaR 0.125): 56.8. Result ID: 6a03328b47f34ce924e7db15. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"alpha (α)":"0.125","objective":"CVaR"}. This record does not assert independent reproduction. ### IPO CVaR α=0.25 Combined Win Rate (CVaR 0.125): 56.5. Result ID: 6a03328b47f34ce924e7db14. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"alpha (α)":"0.25","objective":"CVaR"}. This record does not assert independent reproduction. ### EG k = 1 Neutral Combined Win Rate (CVaR 0.125): 54.9. Result ID: 6a03328b47f34ce924e7db10. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"k":"1","objective":"Neutral"}. This record does not assert independent reproduction. ### IPO Entropic τ=5 Combined Win Rate (CVaR 0.125): 54.8. Result ID: 6a03328b47f34ce924e7db11. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"tau (τ)":"5","objective":"Entropic"}. This record does not assert independent reproduction. ### NMD k = 1 Neutral Combined Win Rate (CVaR 0.125): 52.5. Result ID: 6a03328b47f34ce924e7db0f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"k":"1","objective":"Neutral"}. This record does not assert independent reproduction. ### EGPO† (HF) Combined Win Rate (CVaR 0.125): 11. Result ID: 6a03328b47f34ce924e7db0d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"source":"Hugging Face"}. This record does not assert independent reproduction. ### SFT Base Combined Win Rate (CVaR 0.125): 1.9. Result ID: 6a03328b47f34ce924e7db0c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"type":"Baseline"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.