First of all, I'm not a lab, this was a solo summer research project that finally culminated into the github repo and the writeup. The repo includes a much deeper dive with methods, findings about quantization and geometry, limitations, and proposed experiments. I'll also mention that this is far from production-grade, it's mostly a theoretical framework with a "fake" quantization demo as it lacks real kernels. The geometry of LLMs embedding spaces is highly uneven with a few feature...