What determine the order of the runs in TensorBoard's embeddings tab?

Some examples:

Solution

The TensorBoard embedding project backend serves the runs as a list of keys in a Python dict object:

@wrappers.Request.application
def _serve_runs(self, request):
  """Returns a list of runs that have embeddings."""
  return Respond(request, list(self.configs.keys()), 'application/json')

Quoting the Python documentation for Dictionaries:

Performing list(d.keys()) on a dictionary returns a list of all the keys used in the dictionary, in arbitrary order (if you want it sorted, just use sorted(d.keys()) instead).