I don't like the "motivation-as-introduction" section as it does not present a wide scope within which the paper lies.

All of the reviewers found the proposed technique original and the theory interesting. The reviewers initially had concerns regarding the structure of the paper, relevance of some of the experiments, and comparison with perceptual loss. These concerns are alleviated given the author feedback. Assuming that the authors will integrate the author feedback into the paper and incorporate all of reviewers' feedback, I recommend acceptance as a poster.

Given a trained neural network, we aim at understanding how similar it considers any two samples. For this, we express a proper definition of similarity from the neural network perspective (i.e. We study the mathematical properties of this similarity measure, and show how to estimate sample density with it, in low complexity, enabling new types of statistical analysis for neural networks. We also propose to use it during training, to enforce that examples known to be similar should also be seen as similar by the network. We then study the self-denoising phenomenon encountered in regression tasks when training neural networks on datasets with noisy labels.

Given a trained neural network, we aim at understanding how similar it considers any two samples. For this, we express a proper definition of similarity from the neural network perspective (i.e. We study the mathematical properties of this similarity measure, and show how to estimate sample density with it, in low complexity, enabling new types of statistical analysis for neural networks. We also propose to use it during training, to enforce that examples known to be similar should also be seen as similar by the network. We then study the self-denoising phenomenon encountered in regression tasks when training neural networks on datasets with noisy labels.

The results in figure 6 show the average and standard deviation over 60 runs for each curve. Similar results were observed on all cases. More specifically, the model is made out of 4 neural networks. The resulting amount of similarities to compute would be around half a billion. We thus obtain 3045 patches representing the dataset.