x
Uncertainty in Deep Learning — LessWrong