> [!NOTE]
> Let $\tau: V \times W \rightarrow K$ be a [[Bilinearity|bilinear map]]. Then the $n \times m$ matrix $A=\left(\tau(\mathbf{e}_{i}, \mathbf{f}_{j})\right)$ is defined to be the matrix of $\tau$ with respect to the bases $\mathbf{e}_1, \ldots, \mathbf{e}_n$ and $\mathbf{f}_1, \ldots, \mathbf{f}_m$ of $V$ and $W$.
**Compatibility with matrix multiplication**: For $\mathbf{v} \in V, \mathbf{w} \in W$, let $\mathbf{v}=x_1 \mathbf{e}_1+\cdots+x_n \mathbf{e}_n$ and $\mathbf{w}=y_1 \mathbf{f}_1+\cdots+y_m \mathbf{f}_m$, so the coordinates of $\mathbf{v}$ and $\mathbf{w}$ with respect to our bases are
$
\underline{\mathbf{v}}=\left(\begin{array}{r}
x_1 \\
x_2 \\
\cdot \\
\cdot \\
x_n
\end{array}\right) \in K^{n, 1}, \quad \text { and } \quad \underline{\mathbf{w}}=\left(\begin{array}{r}
y_1 \\
y_2 \\
\cdot \\
\cdot \\
y_m
\end{array}\right) \in K^{m, 1}
$
Then it follows from the defiitomn of $\tau$ that
$
\tau(\mathbf{v}, \mathbf{w})=\sum_{i=1}^n \sum_{j=1}^m x_i \tau\left(\mathbf{e}_i, \mathbf{f}_j\right) y_j=\underline{\mathbf{v}}^{\mathrm{T}} A \underline{\mathbf{w}} .
$
So once we've fixed bases of $V$ and $W$, every bilinear map on $V$ and $W$ corresponds to an $n \times m$ matrix, and conversely every matrix determines a bilinear map.