Viewing profile — Garpagan

> I don't understand how you're so caught up on the term "distillation". Distillation is using a larger model's outputs to train a (weaker) student model. Which is exactly what's h…

There are few problems that I have with the distillation attack claim, and especially using model's self identification as a proof: 1. Kimi's model is almost on par with the SOTA m…

I was reading the article and it made me think. I read through the comments here, and yours was resonating a bit with what was on my mind. So, I apologize for butting in. The artic…