Thanks for the insight. So, I suppose the follow-up question would be whether the techniques of the paper would make it possible to reformulate DOT without exp

Thanks for the insight. So, I suppose the follow-up question would be whether the techniques of the paper would make it possible to reformulate DOT without explicit types. Sorry if that's an ignorant question.