Remix.run Logo
theGnuMe a day ago

You can have small epsilons instead of zeros.

ruined a day ago | parent [-]

what, for all possible words?

3eb7988a1663 a day ago | parent [-]

Instead of a naive dense matrix, you can use some implementation that allows sparsity. If element does not exist, gets a non-zero value which can still be sampled. Which theoretically enables all outputs.

ruined 8 hours ago | parent [-]

i think at that point it's definitionally not a markov chain anymore. how do you sample an open set of unknown values?