Add test coverage for edge cases in neuron selection (neuron_ids / indexes)

[AbirajKangotra]

Problem

Hi! I was going through the recent changes around neuron selection (neuron_ids and indexes) and had a quick question regarding test coverage.

From what I saw, the feature introduces some non-trivial logic (like mutual exclusivity and different selection paths), but I wasn’t sure if edge cases are currently covered by tests.

For example, in the selection logic there is handling like:

if neuron_ids is not None and indexes is not None:
raise ValueError("Only one of neuron_ids or indexes can be provided")

and selection based on inputs:

if neuron_ids is not None:
# select based on neuron_ids
elif indexes is not None:
# select based on indexes

But I couldn’t find clear tests covering cases like:

• Passing both neuron_ids and indexes
• Invalid or non-existent neuron_ids
• Out-of-range indexes
• Empty or duplicate inputs

---

Why I’m bringing this up

These cases might lead to unexpected behavior if not explicitly tested, especially as the feature evolves.

---

Question

Does it make sense to add tests for these edge cases?
If yes, I’d be happy to work on this and open a PR.

---

Let me know if this aligns with the current direction — just wanted to confirm before starting.