It can happen on most LLMs and is usually programmed to decentivize repeating text heavily.

I believe what happens is that when the LLM is choosing what word to use, it looks back on the sentence and sees that it talked about knives, so it wants to continue talking about knives, then it gets itself into a loop.

Scott Pilgrim vs. The World.

Great movie.

This is the same as me. Outside of work, though, I’m the cat in the meme.

Thunder has keyword filters. These help but aren’t perfect.

I switched to using CBC’s Mauril from Duolingo, and it’s been good for me. However, it’s only available to Canadian residents (or VPN users) and it’s only for French. So it worked for me but obviously that covers a small subset of Duolingo users.

It’s posting dumb and usually low effort memes or jokes that aren’t meant to be taken too seriously.