Yes:
float[] values = {0.1f, 0.2f, 0.3f};
This syntax is only permissible in an initializer. You cannot use it in an assignment, where the following is the best you can do:
values = new float[3];
or
values = new float[] {0.1f, 0.2f, 0.3f};
Trying to find a reference in the language spec for this, but it’s as unreadable as ever. Anyone else find one?