#neural-language-model

A neural language model looks up 50-dimensional embeddings for the previous 3 words, concatenates them, applies a hidden layer $g(\mathbf{x}\mathbf{W}^1 + \mathbf{b}^1)$ with 100 units, and then a softmax layer over a vocabulary of 10,000 words. How many parameters do $\mathbf{W}^1$ and $\mathbf{b}^1$ have together?