Theory of Mind reasoning on Hi-ToM
100Accuracy (0th-Order)CoT
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CoTModel=GPT-5.42026.06 | 100 | 88.75 | 60 | 68.75 | 71.25 | 77.75 | |
| SIMTOMModel=GPT-5.42026.06 | 100 | 92.5 | 88.75 | 87.5 | 80 | 89.75 | |
| TIMETOMModel=GPT-5.42026.06 | 100 | 92.5 | 90 | 87.5 | 81.25 | 90.25 | |
| RECTOMModel=GPT-5.42026.06 | 100 | 100 | 100 | 100 | 100 | 100 | |
| CoTModel=Gemini-32026.06 | 100 | 100 | 92.5 | 85 | 83.75 | 92.25 | |
| SIMTOMModel=Gemini-32026.06 | 100 | 91.25 | 90 | 88.75 | 86.25 | 91.25 | |
| TIMETOMModel=Gemini-32026.06 | 100 | 78.75 | 86.25 | 86.25 | 87.5 | 87.75 | |
| RECTOMModel=Gemini-32026.06 | 100 | 100 | 97.5 | 97.5 | 97.5 | 98.5 | |
| CoTModel=Qwen3.52026.06 | 100 | 95 | 88.75 | 83.75 | 85 | 90.5 | |
| SIMTOMModel=Qwen3.52026.06 | 100 | 87.5 | 90 | 90 | 86.25 | 90.75 | |
| TIMETOMModel=Qwen3.52026.06 | 100 | 70 | 87.5 | 86.25 | 82.5 | 85.25 | |
| RECTOMModel=Qwen3.52026.06 | 100 | 100 | 100 | 100 | 100 | 100 | |
| CoTModel=Gemma-42026.06 | 100 | 93.75 | 65 | 62.5 | 58.75 | 76 | |
| RECTOMModel=Gemma-42026.06 | 100 | 100 | 100 | 100 | 97.5 | 99.5 | |
| SIMTOMModel=Gemma-42026.06 | 98.75 | 91.25 | 81.25 | 73.75 | 73.75 | 83.75 | |
| TIMETOMModel=Gemma-42026.06 | 97.5 | 70 | 60 | 63.75 | 60 | 70.25 |