Actually, the probabilities being negative is a problem, because this forces everyone to write a wrong solution.
You can notice that there should be no need to multiply by box_confidence when doing argmax, because you are just multiplying all the compared numbers by the same value. But because the test values wind up negative, this multiplication flips the sign of box_class_probs and the minimum becomes maximum.
Reposting from Box_confidence and box_class_probs - #5 by faucct