“…The relevant question here is to distinguish between two scenarios: the first, in which small interactions are not able to generate new phases and their main effect is a shift of the critical curves, as found in certain 3D topological insulators [45][46][47] ; and a second one, characterized by the emergence of a novel, interaction-induced, topological phase, like the one corresponding to σ 12 = ± e 2 h , predicted for the HaldaneHubbard model in Ref. [26][27][28][29]39,40 . Regarding the longitudinal conductivity, in the absence of interactions it is equal to , for all the values of (φ, W ) on the critical lines, with the exception of the crossing points (φ, W ) = (0, 0), (π, 0), where it is equal to e 2 h π 2 (of course, away from the critical lines σ ii = 0).…”