Abstract: The learning from intervention (LfI) approach has been proven effective in improving the performance of RL algorithms; nevertheless, existing methodologies in this domain tend to operate ...