Agent Task Completion on ToolSandbox
0.67Average Task RewardGPT-5.1 with H-EPM
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.1 with H-EPMModel=GPT-5.1, Enhancement=H-EPM2025.12 | 0.67 | |
| GPT-5.1 baseModel=GPT-5.1, Configuration=Base2025.12 | 0.647 |
| Method | Links | |
|---|---|---|
| GPT-5.1 with H-EPMModel=GPT-5.1, Enhancement=H-EPM2025.12 | 0.67 | |
| GPT-5.1 baseModel=GPT-5.1, Configuration=Base2025.12 | 0.647 |