CV suggesting taxa not known from the region with confidence

General
1

Is it just me or has the iNat CV been suggesting a lot of taxa which aren’t known from the region recently with confidence. It’s gotten to the point that I have to review every ‘confident’ suggestion to make sure it’s known from the region and if it resembles the taxa I am trying to upload correctly. The problem has only come up recently, as before, the frequency of suggestions that were not known from the region was very few, and if present, suggested without confidence. Can anyone on the iNat team please confirm if this is a bug, and if this would be fixed soon? I believe this might already be creating havoc for identifiers.

2

do you have any screenshots or examples? Difficult to diagnose any issues without some more details

2 Likes
3

Did you read this thread when the issue first came up ?

https://forum.inaturalist.org/t/cv-suggestions-have-gotten-much-much-worse-at-san-jacinto-mountain/68136

The 2 subsequent CV updates reverted to the previous

With the model 2.22 update, we changed geomodel approaches from the GRID approach we had been using to a slightly different SINR approach that we thought would be an improvement. However, due to some issues surfaced by community members, we’ve transitioned back to GRID for 2.24 until we are more confident in the alternative approach.

Yesterday’s CV update https://www.inaturalist.org/blog/118979

4

May be relevant:

Species not expected nearby are being presented as “expected nearby”