NeurIPS 2019
Sun Dec 8th through Sat the 14th, 2019 at Vancouver Convention Center
Paper ID:4102
Title:Latent Weights Do Not Exist: Rethinking Binarized Neural Network Optimization


		
This paper proposed a new training method for neural networks with binary weights. The main idea is to not use the existing "latent weights approach" which treats the weights as continuous, rather a new method that relies on the sign of the weights. The proposed approach is based on momentum. Before rebuttal, the authors found the paper to be original, novel, and also simpler than existing methods. They had some concerns regarding the experiments and also a few other small concerns. Post rebuttal, their views changed and they have increased their scores. One main reason is due to the ImageNet result. However, some other concerns remains and authors are encouraged to resolve those in the final version of the paper if accepted. In its current state, I think the paper can be a good addition to NeurIPS. Therefore, I vote for an acceptance.