See https://github.com/JuliaReinforcementLearning/ReinforcementLearning.jl/issues/1067 Such mistakes should throw an error when inserting incorrectly sized traces.