Game playing on Truth and Deception
84Win RateMulti-Game Model
Evaluation Results
| Method | Links | |
|---|---|---|
| Multi-Game ModelOpponent=Gemini-2.0-Flash2025.06 | 84 | |
| Simple Negotiation SpecialistOpponent=Gemini-2.0-Flash2025.06 | 82.8 | |
| Kuhn Poker SpecialistOpponent=Gemini-2.0-Flash2025.06 | 81.6 | |
| TicTacToe SpecialistOpponent=Gemini-2.0-Flash2025.06 | 79.9 | |
| Instruct ModelOpponent=Gemini-2.0-Flash2025.06 | 75.5 | |
| Base ModelOpponent=Gemini-2.0-Flash2025.06 | 49.6 |