Why is the algorithm soo different in the scene version that the message with a single character version? I have been testing Deep Squeak and in the Scene's it is hilarious and spot on but sometimes it gets too in its head. It matches characters that it searches from the Web spectacularly but it is also missing that empath or Heart Warming note.
In the DM - after trying the scenes I find it a bit boring 😄 but fair enough since the Scene version is soo hilarious. It is still good but the difference is like huge.
The only issue with the Scene Specific Style (at least the one I tested - DeepSqueak) was that there was a lot of humor and even when I tried to not be humor and hard core it kept diving there. I do prefer it over the normal one tho - don't get me wrong but sometimes it goes over the top no matter how many swipes or Negative Feedback I give it. On the plus side it includes multiple characters (world building) but sometimes it is as if it ignores its own messages and mine with that, like it says something completely opposite of what it said earlier. Whereasa in the DM it is more one on one and getting multiple characters in a single response requires a bit of work.
But legit, I feel like it should be split into two different styles, like if it is the scenario itself that makes the Algorithm act that way with parameters we set we should be able to save it so we can swap between.
FYI I would have opened an issue on https://support.character.ai/hc/en-us/requests/new but it doesn't seem to be working.