“…Earlier works 38,48,49 had poined out that positive exchange bias results from competition between antiferromagnetic exchange coupling across ferromagnetic/antiferromagnetic interface (i.e., negative J int ) and ferromagnetic coupling between applied field and antiferromagnetic moment at the interface. However, as shown in our earlier theoretical work 30 , neither in two-dimensional nor in threedimensional spin structure, variation of J int (including switch in J int from positive to negative) has any significant impact on the exchange bias for Ising, XY, and Heisenberg spins. The role of J int in governing the sign and magnitude of exchange bias could be significant in nanoscale core-shell structures where the core is ferromagnetic and the shell is antiferromagnetic and the measurement is carried out, primarily, under field-cooled condition.…”