Traces
Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.
Snooker Prediction
task page →5 runs · indexed 2026-09-02| # | Model | Reward | Cost | In tok | Out tok | Time | Steps | Submits |
|---|
| 1 | GPT-5.6 | 0.4857 | $119.18 | 111.7M | 242k | 2h 5m | 299 | 11 |
| 2 | GPT-5.6 | 0.4839 | $170.16 | 165.0M | 275k | 4h 12m | 432 | 9 |
| 3 | GPT-5.6 | 0.4428 | $157.87 | 150.8M | 208k | 2h 34m | 354 | 9 |
| 4 | GPT-5.6 | 0.4295 | $132.32 | 126.4M | 240k | 4h 1m | 369 | 6 |
| 5 | GPT-5.6 | 0.2890 | $179.27 | 154.6M | 213k | 9h 23m | 357 | 5 |