velcroscientist

4 ms·

1. 1.▲
   by velcroscientist 5y ago

   Good point. This is why robotics researchers do not take deep RL papers seriously unless they have some real world robotics results. I'm looking at you, people who only show mujoco results and claim their algorithm is useful for roboti
2. 2.▲
   A satirical comic on the cult of deep learning [pdf](github.com)

   3 points by velcroscientist 6y ago|0 comments
3. 3.▲
   by velcroscientist 7y ago

   You're correct. As long as your test set remains within the training distribution, you can expect the NN to be well behaved. However, its behavior is undefined for testing data out of training distribution. There is a lot of work on de
4. 4.▲
   OpenAI's GPT2 Model as Slack Integration(github.com)

   6 points by velcroscientist 8y ago|0 comments