Not necessarily. In free fall it will be stable with either the heavy or the light side toward the Earth. Whichever is lower will accelerate faster, thus tending to align the long axis of the mass distribution with the gravity vector. That is different from the case of an unbalanced wheel on a stationary bearing. In that case the bearing places a nongravitational constraint on the motion, a constraint that is not present with a satellite in free fall.

Clear as mud? I realize this can be hard to visualize from words alone.