1 paper · 1 filter
Chong He, Yuechen Luo, Fang Li +2
End-to-end autonomous driving models often encounter performance bottlenecks, as training-time scaling leads to high computational costs and diminishing marginal returns. Existing…