Troop placement prediction on Risk
0.51EMDMCTS with Const-o-T
Evaluation Results
| Method | Links | |
|---|---|---|
| MCTS with Const-o-TModel=GPT-4.12025.10 | 0.51 | |
| CoTModel=GPT-4.12025.10 | 0.54 | |
| MCTS with Const-o-TModel=GPT-5.22025.10 | 0.54 | |
| MCTS with Const-o-TModel=Claude Haiku2025.10 | 0.55 | |
| CoTModel=GPT-5.22025.10 | 0.56 | |
| CTPModel=GPT-4.12025.10 | 0.56 | |
| CoVModel=GPT-4.12025.10 | 0.56 | |
| MCTS with Const-o-TModel=Llama 3.32025.10 | 0.57 | |
| Direct PromptModel=GPT-5.22025.10 | 0.58 | |
| CTPModel=GPT-5.22025.10 | 0.59 | |
| CoVModel=GPT-5.22025.10 | 0.59 | |
| MCTS with Const-o-TModel=DeepSeek R12025.10 | 0.59 | |
| MCTS with Const-o-TModel=GPT-OSS-120B2025.10 | 0.59 | |
| Direct PromptModel=GPT-4.12025.10 | 0.6 | |
| CTPModel=GPT-OSS-120B2025.10 | 0.6 | |
| Const-o-TModel=GPT-4.12025.10 | 0.6 | |
| Direct PromptModel=GPT-OSS-120B2025.10 | 0.61 | |
| Const-o-TModel=GPT-5.22025.10 | 0.61 | |
| CoVModel=GPT-OSS-120B2025.10 | 0.62 | |
| Direct PromptModel=Claude Haiku2025.10 | 0.63 | |
| CoVModel=Claude Haiku2025.10 | 0.63 | |
| Direct PromptModel=Llama 3.32025.10 | 0.64 | |
| CoTModel=Claude Haiku2025.10 | 0.64 | |
| CoT+RSModel=GPT-4.12025.10 | 0.64 | |
| Const-o-TModel=GPT-OSS-120B2025.10 | 0.64 | |
| CTPModel=Claude Haiku2025.10 | 0.65 | |
| Const-o-TModel=Claude Haiku2025.10 | 0.65 | |
| CoT+RSModel=GPT-OSS-120B2025.10 | 0.66 | |
| Const-o-TModel=Llama 3.32025.10 | 0.66 | |
| CTPModel=Llama 3.32025.10 | 0.68 | |
| CoTModel=GPT-OSS-120B2025.10 | 0.7 | |
| CoT+RSModel=GPT-5.22025.10 | 0.7 | |
| CoT+RSModel=Claude Haiku2025.10 | 0.7 | |
| CoVModel=Llama 3.32025.10 | 0.7 | |
| MCTS with CoTModel=GPT-5.22025.10 | 0.7 | |
| CoT+RSModel=Llama 3.32025.10 | 0.72 | |
| ToTModel=GPT-4.12025.10 | 0.72 | |
| CoTModel=DeepSeek R12025.10 | 0.73 | |
| CoTModel=Llama 3.32025.10 | 0.74 | |
| MCTS with CoTModel=GPT-4.12025.10 | 0.75 | |
| Direct PromptModel=DeepSeek R12025.10 | 0.76 | |
| ToTModel=GPT-5.22025.10 | 0.76 | |
| MCTSModel=GPT-4.12025.10 | 0.76 | |
| MCTS with CoTModel=Claude Haiku2025.10 | 0.76 | |
| Const-o-TModel=DeepSeek R12025.10 | 0.77 | |
| CoVModel=DeepSeek R12025.10 | 0.79 | |
| MCTSModel=GPT-5.22025.10 | 0.79 | |
| MCTSModel=Claude Haiku2025.10 | 0.79 | |
| LLMFPModel=GPT-4.12025.10 | 0.81 | |
| LLMFPModel=GPT-OSS-120B2025.10 | 0.82 | |
| MCTS with CoTModel=GPT-OSS-120B2025.10 | 0.82 | |
| CTPModel=DeepSeek R12025.10 | 0.85 | |
| ToTModel=GPT-OSS-120B2025.10 | 0.85 | |
| CoT+RSModel=DeepSeek R12025.10 | 0.87 | |
| ToTModel=Claude Haiku2025.10 | 0.89 | |
| ToTModel=DeepSeek R12025.10 | 0.9 | |
| MCTSModel=GPT-OSS-120B2025.10 | 0.93 | |
| MCTSModel=Llama 3.32025.10 | 0.96 | |
| MCTS with CoTModel=Llama 3.32025.10 | 0.99 | |
| MCTSModel=DeepSeek R12025.10 | 1.01 | |
| MCTS with CoTModel=DeepSeek R12025.10 | 1.01 | |
| LLMFPModel=Claude Haiku2025.10 | 1.03 | |
| ToTModel=Llama 3.32025.10 | 1.05 | |
| LLMFPModel=GPT-5.22025.10 | 1.08 | |
| LLMFPModel=Llama 3.32025.10 | 1.13 | |
| LLMFPModel=DeepSeek R12025.10 | 1.15 |