Same-category sample in the negative queue

Thank you for effort on these papers and the code.
I have some questions on sampling negative matches.

1) the queue contains vectors of previous images, and will be used as negative samples later, but we do not know if the vectors come from an image of a same category. i.e., we are working on a img_q which is a dog, but in the queue, there could be a lot of dog-vectors from previous images, and now they are regarded as negative to any new images. May I know if the situation exist in MOCO?
2) If the situation does exist, may I know if it is alleviated by the large negative sample size (i.e. 65535)?

Looking forward to your reply!