Loss‑Band Sparsity Assumption in Scientist AI Fails Under Empirical Test
An empirical probe of the loss‑band sparsity assumption behind Scientist AI finds the volume claim holds while the curvature claim does not. Using one model and subspace, the author shows the dangerous‑set fraction is low in low‑loss bands, and highlights the localized undirected perturbation methodology as the key contribution, casting doubt on the safety guarantee.