I've been using n-grams heavily in my PhD. I like a lot what you are describing. Also a simple extension is to use "cache" n-gram models, that can naively simul

I've been using n-grams heavily in my PhD. I like a lot what you are describing. Also a simple extension is to use "cache" n-gram models, that can naively simulate longer-term "memory". What's fascinating is that in the entropy/information-theoretic sense, instead of using wpm, one could aim for "constant" information rate of reading (as in bits/token). That's really cool as a concept. I don't know if it would work, obviously. I'd definitely be up to for hacking this up, but currently a bit too busy. Will ping you at some point though!