BreakingWeb/breakingweb-results-v3 download history
BreakingWeb/breakingweb-results-v3 is a reinforcement learning dataset on the Hugging Face Hub. In the last 30 days it was downloaded 620 times (163 in the last 7 days), and 620 times in total. It ranks #27,872 among datasets by monthly downloads.
BreakingWeb — agent trajectories, part 2: Sonnet 4.6, Opus 4.7 retry pass, pixel agents Part of the BreakingWeb benchmark release: 519 matched clean / intervention browser-task pairs across 7 self-hosted web environments, scored against live backend state. This repo completes the pape
Open BreakingWeb/breakingweb-results-v3 on Hugging Face
Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.