I presented our recent paper on correcting RL policies using non-expert feedback at HRI 2022. You can watch it here: HRI 2022 Talk.