import public Data.Tensor.Softargmaxsoftargmax : Fractional a => Exp a => Ord a => Neg a => IsFoldable (i .cont) => AllAlgebra [i] a => Tensor [i] a -\-> Tensor [i] aSoftargmax as a parametric map, with temperature as a parameter
The underlying implementation lives in `Data.Tensor.Softargmax`
TODO since distribution is an applicative functor (https://glaive-research.org/2025/02/11/Generalized-Transformers-from-Applicative-Functors.html)
is there a meaningful notion of the "distribution container"?
Is there a sense in which `Dist` is a functor on containers?