Further testing the validity of generalized heterogeneous-elasticity theory for low-frequency excitations in structural glasses

In a recent paper E. Lerner and E. Bouchbinder, Phys. Rev. E {\bf 111}, L013402 (2025) raised concerns regarding the validity of the theory and the interpretation of the data, presented in our previous study on non-phononic vibrational excitations in glasses, W. Schirmacher {\it et al.}, Nature Comm. {\bf 15}, 3107 (2024). In that work, we presented evidence suggesting that the commonly observed low-frequency regime of the non-phononic vibrational density of states (DoS) is, in fact, highly dependent on technical aspects of the molecular dynamics simulations employed to compute the DoS. Specifically, we showed that the $ω^4$ scaling results from the use of a tapering function applied to ensure continuity of the interaction potential at the cutoff distance. In this letter we report further evidence in favor of our previous findings, namely the non-universality of the $ω^4$ DoS scaling, and the existence of an important class of non-phononic excitations in glasses, which we call defect states, and are induced by frozen-in stresses. These modes can be classified as quasi-localized.