Tensor
N-dimensional numeric array; the core ML data structure.
Tensor is the basic data structure of modern machine learning
: a grid of numbers with
any number of dimensions. It generalizes the familiar cases (a single number is a 0-D tensor,
a list of numbers is 1-D, a table (matrix) is 2-D) to as many dimensions as you need; a batch
of color images, for instance, is a 4-D tensor of shape [batch, height, width, channels].
Every input, weight, and intermediate value in a neural network is a tensor, and training is basically a long chain of tensor operations (matrix multiplies, additions, and simple nonlinear steps). Frameworks like PyTorch, TensorFlow, JAX, and MLX are built around tensors and run those operations on accelerators (GPUs/TPUs), hardware whose whole job is doing this kind of math massively in parallel.
Two properties matter most day to day: shape (the size along each dimension, the usual source of bugs) and dtype (how precisely each number is stored, e.g. fp32/fp16/bf16, trading accuracy for speed and memory). One note: the ML “tensor” is really just a multidimensional array, and is looser than the stricter math/physics meaning of the word.