Traces

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

Granite Mamba2 Inference Optimization

task page →5 runs · indexed 2026-09-02
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Gemini 3.7 Flash0.7839$8.9784.4M442k56m2813
2Gemini 3.7 Flash0.3912$15.89144.8M798k2h 17m4795
3Gemini 3.7 Flash0.3149$10.7497.9M514k1h 12m2782
4Gemini 3.7 Flash0.2425$12.51121.4M559k2h 10m3262
5Gemini 3.7 Flash0.2101$6.7049.8M379k47m2123