May be these comments are actually originating from training annotated data? If I were to add code annotations for training data, I would sort of expect such comments which makes not much value for me but for the model, gives more contextual understanding…