Nicholas Goldowsky-Dill comments on Detecting Strategic Deception Using Linear Probes