Non-uniform circuits, according to my understanding, are those which have different circuit depending on the input size. Constant depth circuit are those whose depth is constant in the input size. So if for example we considered an instance k of the complexity class TC0 which is a non-uniform constant-depth circuit, and let kn be the circuit instance at input size n, does that mean that the following sentence is allowed or not: depth(ki) ≠ depth(kj) when i≠j.
In other words, does the non-uniformity of a complexity class allow a constant-depth circuit to have different depths for different input size or not ?

