Regarding the sum of softmax values exceeding 1.0

In a 1000 class classification task, when using the softmax function, Top1 is 1.0 and Top2 is greater than or equal to 0. This makes the total more than 1.0. Does anyone know anything about this?
The environment is tensorflow2.6.1 and python3.8.10.

Top1 - score:1.00000000000000000000000000000000
Top2 - score:0.00000002750548944163710984867066
Top3 - score:0.00000000408126732409641590493266
Top4 - score:0.00000000309478398463625126169063
Top5 - score:0.00000000193547711191399685048964


Welcome to the Tensorflow Forum!

This is unintended behavior. Could you please provide a standalone code to replicate the issue reported here?

Thank you!