why TrainableQuanvFilter output float instead 01？

[lianghaichen]

when i print the output of function forward of class TrainableQuanvFilter, it shows as follow.in my opinion,after code "data = self.measure(qdev)",the output is a list of 0 or 1(state of a qbit)，why output is float?

([[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619],
[-0.5682, -0.2191, 0.2454, -0.0619]], device='cuda:0',
grad_fn=)