Dialogue Goal Completion on Persona-Scenario
9.19Believability (Sotopia-Eval)Resonant Minds
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Resonant MindsInformation Access=Realistic information asymmetry, Statistic=Mean ± Std, Number of evaluation runs=32026.06 | 9.19 | 7.92 | -1.01 | -0.04 | 88.23 | 97.13 | 94.25 | 94.98 | |
| ScriptInformation Access=Full-information reference (unrealistic), Statistic=Mean ± Std, Number of evaluation runs=32026.06 | 8.95 | 8.87 | -2.33 | -0.09 | 89.49 | 97.65 | 94.72 | 93.26 | |
| AgentInformation Access=Realistic information asymmetry, Statistic=Mean ± Std, Number of evaluation runs=32026.06 | 8.17 | 6.54 | -1.77 | -0.07 | 82.98 | 98.11 | 89.11 | 88.29 |