b) In the approach “Curiosity-driven Exploration by Self-supervised Prediction”, in environments where all states are equally novel or predictable, this curiosity-driven approach would behave identically to a random exploration strategy.