[adorable]

I guess you are referring to http://cs.nyu.edu/~zaremba/docs/understanding.pdf

For those interested, the papers and discussion are part of this weekly collection of AI-related news and resources: https://aiweekly.curated.co

[Houshalter]

He's also referring to this follow up paper which attempts to explain the issue and show that it exists in other models too: http://arxiv.org/abs/1412.6572