Feature importance across wavelengths for all parameters included in our generated model grid for the case study of the Y1 dwarf WISEPAJ1541-22 as well as their joint retrieval. The significance of a feature is quantified by the normalized decrease in variance it generates throughout the training process. Specifically, it denotes the total reduction in variance achieved whenever that feature is employed for partitioning nodes in a decision tree within the ensemble. For visual guidance, the rescaled spectrum of WISEPAJ1541-22 is overlaid as solid black lines.