that put the Capital of Denmark on the left of the rest of the country. Overall the model shows signs of what we could interpret as a "resolution limit" . I'm sure this limit is different in different countries and it's probably connected with the training data, but it's something worth remembering.

In many cases - mostly the smaller parties - the actual % of votes is quite within the range of the predicted values. Yet, the model gets wrong the top party, over estimates the Socialdemocratic party and under estimates Enhedslisten. These are all - well-known - specific traits of Copenhagen,

(multiple iterations are shown) as well as the actual % of votes. The results are somehow interesting and shows the potential and the limitations of this approach. Here I'm going to refer to the "values" produced by the LLMs as "predictions" (bear with me).

Michelle has been invaluable in that role for nearly a decade now, but the time has come for @aoir.bsky.social to find a new coordinator to take on these vital duties. 2/2

Job posting is on our listserv. Please share & apply! listserv.aoir.org/pipermail/ai...