08-06-2025, 05:19 PM
Interesting work. I agree with you (and many others) that the structure of Voynich word types is rather regular, with some underlying 'grammar', and I think the method you used is surely interesting.
I had my share of fun working on grammars (actual slot grammars, in my case). Now I don't have the time at hand to check your GitHub repository and see what I can understand, I confess I did not understand much of how your method works and what it actually does, but I'll try tomorrow.
One question: can you apply your method in 'reverse', I mean: use your GPT model to decompose the original Voynich word types in parts ('tokens'), such as "qo", "y", "dy", "aiin" and so on, for each word (some of which may turn out not to be conforming to the grammar)? In this case it could be possible to compare the results with different approaches (I defined a metric which ranks different grammars according to how many bits of information are needed to represent the text by using optimal encoding on the words divided in 'tokens').
I had my share of fun working on grammars (actual slot grammars, in my case). Now I don't have the time at hand to check your GitHub repository and see what I can understand, I confess I did not understand much of how your method works and what it actually does, but I'll try tomorrow.
One question: can you apply your method in 'reverse', I mean: use your GPT model to decompose the original Voynich word types in parts ('tokens'), such as "qo", "y", "dy", "aiin" and so on, for each word (some of which may turn out not to be conforming to the grammar)? In this case it could be possible to compare the results with different approaches (I defined a metric which ranks different grammars according to how many bits of information are needed to represent the text by using optimal encoding on the words divided in 'tokens').