Here's how the Feature Importance bar graph looks like:

Typically, Decanter AI would compute each feature's explainability against the trained model and display a bar graph with the feature's importance ranked from highest to lowest for each model. However, Stacked Ensemble's models are trained by multiple algorithms rather than just one particular algorithm, therefore is it not possible to compute feature importance against one model's explainability. 

Tip: If feature importance ranking is required for your project, you can refer to the best-trained model's feature importance bar graph for this purpose.