Inner most dimension of an Array

Question

Which dimension of the array is called the innermost dimension of the array? I was going through some tensorflow documentation "ctc loss" where they describe one jargon called innermost dimension of the array. According to me, there can be three possibilities for the answer. 1. leftmost if see from the right. 2. rightmost if we see from the left. 3. Middle, if we see from both side.

Can someone please explain what innermost dimension means here?. Thanks for going through my doubt.

Answer 1

If we refer to the tensorflow repo, we can get the answer:

// The order of entries in "dim" matters: It indicates the layout of the
// values in the tensor in-memory representation.
//
// The first entry in "dim" is the outermost dimension used to layout the
// values, the last entry is the innermost dimension. This matches the
// in-memory layout of RowMajor Eigen tensors.

(emphasis mine)

This is the same as the default (row-major, also called C-style) layout of numpy arrays, where the last dimension is considered the innermost because it varies the fastest.