Results of normalisation of six synthetic spectra multiplied by six manually fitted pseudo-continua trained with the application of active learning (synthetic data supplemented with manually normalised spectra). On the left of each row, we show the differences between automatically normalised spectra and synthetic spectra, and on the right we show the histograms of those differences with a related spectral type, median with 15.87 percentile in the upper index, and 84.13 percentile in the lower index. The dashed lines on each panel correspond to the residuals, which are equal to −0.01, 0.0, and 0.01, respectively. The use of active learning resulted in a slight reduction in the dispersion of residuals.