I am Jianyang Gao, first author of the RaBitQ papers. I am posting this here because TurboQuant is now being discussed in `r/LocalLLaMA` in the context of local inference / KV-cache compression, and I think the community should have a technically precise comparison on the public record. We are posti
