R1, R6: Additional analyses/ablations for L sparse and L

Neural Information Processing Systems 

We thank the reviewers for their thoughtful comments and suggestions. Below, we address the reviewers' comments individually. We will add these analyses to the main text. Keypoints can indeed "jump" between frames, but we show in a new analysis (Fig. D) that the Jumping thus seems to be a minor issue. R1: What is the size of the feature vector in CNN-VRNN?