Understanding Synthetic Gradients and Decoupled Neural Interfaces · HackerPapers discussions

Understanding Synthetic Gradients and Decoupled Neural Interfaces — mentioned by gwern

Sat, 27 May 2017 16:24:39 GMT

There might be some hard to release infrastructure code for the MCTS part, certainly, but the model on its own should be a standard TF CNN model and highly competitive (and people can write their own MCTS wrapper, it's not that complex an algorithm). Nothing in the AG paper or statements since has hinted at using anything as exotic as synthetic gradients* and there is no reason to use synthetic gradients in AG. (In RL applications the NNs are generally small because there's so little supervision from the rewards so a large NN would overfit grossly; a NN so large as to require synthetic…

Paper and discussion history. A mention is not an endorsement.

Understanding Synthetic Gradients and Decoupled Neural Interfaces — HN submission

Thu, 25 May 2017 17:33:18 GMT

Understanding Synthetic Gradients and Decoupled Neural Interfaces

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Understanding Synthetic Gradients and Decoupled Neural Interfaces — HN submission

Tue, 07 Mar 2017 22:01:04 GMT

Understanding Synthetic Gradients and Decoupled Neural Interfaces

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Understanding Synthetic Gradients and Decoupled Neural Interfaces — HN submission

Sat, 04 Mar 2017 07:06:48 GMT

Deepmind – Understanding Synthetic Gradients and Decoupled Neural Interfaces

6 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Understanding Synthetic Gradients and Decoupled Neural Interfaces — HN submission

Fri, 03 Mar 2017 18:30:34 GMT

Understanding Synthetic Gradients and Decoupled Neural Interfaces

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.