This is why IEEE floating point has -0.0, which represents sequences approaching 0 (underflow) from the negative side; 0.0 represents underflow from the positiv

This is why IEEE floating point has -0.0, which represents sequences approaching 0 (underflow) from the negative side; 0.0 represents underflow from the positive side.