Adding ArrayLike from numpy as a type hint leads to doc build errors.

Description of the problem

I'm trying to improve type hints in the /home/himan/projects/mne-python-issue13597/mne/preprocessing/ica.py file with something like this(one of the functions in that file):

  def get_explained_variance_ratio(
      self,
      inst: BaseRaw | BaseEpochs | Evoked,
      *,
      components: ArrayLike | int | None = None,
      ch_type: Literal["mag", "grad", "planar1", "planar2", "eeg"]
      | ArrayLike
      | None = None,
  ):

however the ArrayLike - sphinx is unable to resolve this ref properly. And give the following error:

 <unknown>:1: WARNING: py:class reference target not found:
  numpy._typing.ArrayLike [ref.class]

Link to data

No response

Expected results

Should build docs without the error described above and add proper ref to the ArrayLike.

Actual results

Throws an error and fails to resolve link.

Additional information

Discussion on this.

Start with the annotations in mne/preprocessing/ica.py and reproduce the documentation build that emits the unresolved numpy._typing.ArrayLike warning. Read the NumPy ArrayLike typing reference and the project’s Sphinx configuration to determine how the reference should resolve; done means the docs build completes without that warning and links ArrayLike correctly.