Traces
Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.
Cranelift Codegen Optimization
task page →5 runs · indexed 2026-09-02| # | Model | Reward | Cost | In tok | Out tok | Time | Steps | Submits |
|---|
| 1 | GPT-5.6 | 0.0176 | $225.59 | 218.6M | 216k | 6h 56m | 506 | 6 |
| 2 | GPT-5.6 | 0.0000 | $297.21 | 276.6M | 238k | 7h 17m | 549 | 5 |
| 3 | GPT-5.6 | 0.0000 | $234.61 | 222.5M | 240k | 7h 38m | 461 | 5 |
| 4 | GPT-5.6 | 0.0000 | $212.19 | 205.4M | 208k | 5h 52m | 461 | 5 |
| 5 | GPT-5.6 | 0.0000 | $355.12 | 344.3M | 254k | 10h 32m | 627 | 3 |