They’ve been thinking about these different dimensions and trying to weave them into the algorithm?
j 下一段next speechk 上一段previous speech
