An Exercise in Sharing: Dabi tells you he wants to try a little something new and you decide to indulge him...

Just a heads up-

1. Multi-character bots, especially ones like this with four personalities, do not work the best on LLM. I tested it, it did OK, but I had to re-roll a decent amount when it would talk for me.

2. Twice is not going to be perfect. His personality alone for me is a lot of information in order to make him speak in his contradictions, and I just cannot fit all that into this bot and have it function. That being said his personality is still there, his speech just won’t be 100% accurate. Every now and then it did work, so you can try to re-roll or fidget with the text if you wanna try.

Please like and comment!

Google Form for Suggestions/Feedback

Tips and Commissions on Ko-Fi

About My Bots

- I test on both OpenAI and LLM.

- I tend to write my bots WITHOUT user description or backstory, to allow you to use your own personas.

- I always ensure that my bots are within the Terms of Service, including aging up any characters. The age is stated in the intro and coding.

-To adhere to the Terms of Service, aged up bots cannot use screenshots. I use AI art through PixAi, as many fan artists do not want their art used for bots.

Info On Bots

This bot keeps talking for me/repeating itself, etc.

AI problem: Sometimes the bot can take over the conversation, it's a common and unsolvable issue. I do my best to manage it on my end. To prevent this, try to avoid short or dry answers that may prompt the bot to take control of the story.

The bot keeps misgendering me, using the wrong names, etc.

AI problem: Utilize chat memory to remind the bot of correct pronouns/gender. I usually write my bots as gender neutral, but mistakes happen. If you notice a gendered term in the intro, leave a comment and I'll fix it. No need for snippy comments.

The bot is being very random, overly sexual, aggressive, etc.

AI problem: Do you think I'm making the bot do these things? Like the bot speaking for you, the AI can sometimes act independently. This is especially true with LLM. Make sure to read the trigger warnings and tags - if it's labeled "Dead Dove" or has a trigger warning for aggression, don't be surprised by the bot's actions.

#dhorrl