Exactly, it's a natural way to write a college essay. I've never not cringed reading an article/blog post that is structured that way, it comes across very contrived. I've also noticed that LLMs tend to prefer it, and humans tend to avoid it in general.
Amazing, I've been wishing for this! Do you have any estimates on how much accuracy is first lost then recovered compared to the original bf16 and the naively quantized models?