Nathan Helm-Burger comments on How well do truth probes generalise?