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 2026-09-22
#ModelRewardCostIn tokOut tokTimeStepsSubmits
1Claude Fable 50.4405$329.75265.7M842k18h 32m62721
2Claude Fable 50.3600$264.94218.2M625k12h 41m43518
3Claude Fable 50.3089$197.41157.5M532k18h 2m36913
4Claude Fable 50.1348$215.61174.0M561k17h 21m40212
5Claude Fable 50.0276$244.97201.6M586k17h 26m45716