V1:

• It's more of a mix of concept and style.

• The trigger words are not actual trigger words, but the most used tags that help to achieve certain aspects.

V2:

• closer to the concept.

• changed some tags, added some tags -> take a look at the "trigger" words or metadata.

V3-Next:

Sorry for the long delay! On top of having very little time, I unfortunately lost my entire dataset.

• The Next version is a preview since it's only a subset of the whole data. The tagging is getting pretty exhaustive, so I made a cut to make sure everything is on the right track—and it feels good to finally share it. :)

• From what I've tested, it feels smoother (thanks to the smaller subset) and much more coherent.