Weighted Finite State Transducers in speech recognition: https://scholar.google.com/scholar?q=finite+state+transducer+speech+recognition&hl=en&as_sdt=0&as_vis=1

Weighted Finite State Transducers in speech recognition: https://scholar.google.com/scholar?q=finite+state+transducer+speech+recognition&hl=en&as_sdt=0&as_vis=1&oi=scholart#d=gs_qabs&t=1693435747591&u=%23p%3DdJXQEQWp23MJ https://scholar.google.com/scholar?q=finite+state+transducer... Modified Kneser-Ney smoothing: https://en.m.wikipedia.org/wiki/Kneser%E2%80%93Ney_smoothing#:~:text=Modified%20Kneser%E2%80%93Ney%20smoothing&text=Computational%20efficiency%20and%20scaling%20to,a%20performant%20open%2Dsource%20implementation https://en.m.wikipedia.org/wiki/Kneser%E2%80%93Ney_smoothing.... We've been here before, neural LMs replaced that generation of models.