Automatic false positive for certain labels (for improving the base model)

[d-dsn]

If I have understood correctly, submitting false positives is a way for the community to help train the base model?

Now what if we could have certain labels to be automatically labeled as "false positives" without the user feedback? I'm looking at the different labels I could choose in Frigate+ and there's 100% chance I will never ever see:
purolator
znpost
horse
raccoon
bear
cow
goat
...

So what if these labels would be also tracked in the background, but always and without questions submitted as false positives.