Question Asking Policy Evaluation on Mixed 20 Question
8.64RVOIVOI
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VOICost=0.01, Model=Gemini-2.5-Flash2026.01 | 8.64 | — | — | 0.34 | 0.54 | |
| VOICost=0.02, Model=Gemini-2.5-Flash2026.01 | 7.72 | — | — | 0.84 | 0.92 | |
| VOICost=0.05, Model=Gemini-2.5-Flash2026.01 | 5.01 | — | — | 1.36 | 1.37 | |
| VOICost=0.10, Model=Gemini-2.5-Flash2026.01 | 1.38 | — | — | -0.9 | 0.48 | |
| VOICost=0.20, Model=Gemini-2.5-Flash2026.01 | -0.96 | — | — | -0.96 | 3.64 | |
| ConfidenceCost=0.01, Model=Gemini-2.5-Flash, Threshold (τ)=0.92026.01 | — | 8.3 | — | — | — | |
| ConfidenceCost=0.02, Model=Gemini-2.5-Flash, Threshold (τ)=0.92026.01 | — | 6.88 | — | — | — | |
| ConfidenceCost=0.02, Model=Gemini-2.5-Flash, Threshold (τ)=0.92026.01 | — | — | 6.8 | — | — | |
| ConfidenceCost=0.05, Model=Gemini-2.5-Flash, Threshold (τ)=0.52026.01 | — | — | 3.64 | — | — | |
| ConfidenceCost=0.10, Model=Gemini-2.5-Flash, Threshold (τ)=0.52026.01 | — | 2.28 | — | — | — | |
| No QuestionCost=0.20, Model=Gemini-2.5-Flash2026.01 | — | 0 | — | — | — | |
| RoundCost=0.01, Model=Gemini-2.5-Flash, Threshold (τ)=152026.01 | — | — | 8.1 | — | — | |
| RoundCost=0.05, Model=Gemini-2.5-Flash, Threshold (τ)=52026.01 | — | 3.65 | — | — | — | |
| RoundCost=0.10, Model=Gemini-2.5-Flash, Threshold (τ)=52026.01 | — | — | 0.9 | — | — | |
| RoundCost=0.20, Model=Gemini-2.5-Flash, Threshold (τ)=52026.01 | — | — | -4.6 | — | — |