Traces

Every run: the agent's full trajectory, its submissions over the 20-hour budget, and the final score.

Multi-GPU Efficient Finetuning

task page →5 runs · indexed 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Grok 4.70.0000$60.2136.3M189k18h 34m1842
2Grok 4.70.0000$37.8924.1M164k18h 10m1402
3Grok 4.70.0000$29.8818.4M159k14h 59m1193
4Grok 4.70.0000$36.1819.8M164k20h 0m1265
5Grok 4.70.0000$59.6535.5M185k17h 44m1703