Update README.md

README.md

@@ -46,10 +46,10 @@ log_weights, log_loss = gradNorm(net=mtlnet, layer=net.fc4, alpha=0.12, dataload
 Consider $T$ regression tasks trained using standard squared loss onto the functions:
 
 $$
-f_i (\mathbf{x}) = \sigma_i  \tanh \left( ( \mathbf{B} + \mathbf{\epsilon}_i ) \mathbf{x} \right)
+f_i (\mathbf{x}) = \sigma_i  \tanh \left( ( \mathbf{B} + \epsilon_i ) \mathbf{x} \right)
 $$
 
-Inputs are dimension 250 and outputs dimension 100, while $\mathbf{B}$ and $\mathbf{\epsilon}_i$ are constant matrices with their elements generated IID from $N(0; 10)$ and $N(0; 3.5)$, respectively. Each task, therefore, shares information in B but also contains task-specific information $\mathbf{\epsilon}_i$. The $\sigma_i$ sets the scales of the outputs.
+Inputs are dimension 250 and outputs dimension 100, while $\mathbf{B}$ and $\epsilon_i$ are constant matrices with their elements generated IID from $N(0; 10)$ and $N(0; 3.5)$, respectively. Each task, therefore, shares information in B but also contains task-specific information $\epsilon_i$. The $\sigma_i$ sets the scales of the outputs.
 
 ```python
 from data import toyDataset