Typo in equivariant chapter (#239)

dl/Equivariant.ipynb

@@ -361,7 +361,7 @@
  "\n",
  "The definition means that we get the same output if we transform the input function to the neural network or transform the output (in the equivariant case). In a specific example, if we rotate the input by 90 degrees, that's the same result as rotating the output by 90 degrees. Take a moment to ensure that matches your idea of what equivariance means. After all this math, we've generalized equivariance to arbitrary spaces and groups.\n",
  "\n",
- "What the two input and output spaces? It's easiest to think about them as the same space for equivariant neural networks. For an invariant, the output space is typically a scalar. Another example for an invariant one could be aligning a molecular structure to a reference. The neural network should align to the same reference regardless of how the input is transformed. \n"
+ "What are the two input and output spaces? It's easiest to think about them as the same space for equivariant neural networks. For an invariant, the output space is typically a scalar. Another example for an invariant one could be aligning a molecular structure to a reference. The neural network should align to the same reference regardless of how the input is transformed. \n"
  ]
  },
  {