The authors regret that the above article used incorrect values in the kMC simulation. The Δg used in the kMC simulation for dislocation nucleation should be the energy differences between the adjacent stable states of each micro-substructures during transition, which abstracted from Fig. 1(b), instead of our previously used energy barriers. The authors have made amendments to the publication as follows: In the section 2.2. Kinetic Monte Carlo study “Δgm,n(0) is the generalized stacking fault energies (GSFEs) at 0 K and stress-free states for the transitions of each micro-substructure, which were directly obtained from Fig. 1(b) and can be seen in Table 3.” “Δek(T)=Δek(1−TTm), where Δek(0) is the stable energy for each transition state at 0 K and stress-free, which is directly obtained from Fig. 1(b) and can be seen in Table 4.” The Table 3 for energy barriers for each transition (Table 3 in the original publication) was moved to Table 5. And the text was revised accordingly. In the section 3. Results “The values of the energy barriers for each transition state in this study can be found in Table 5.” Because the values of Δg used in the kMC simulations were corrected, the kMC simulations (Figs. 4, 5 and 6 in the original publication) using the corrected values were re-performed and newly obtained results are shown in this corrigendum as follows. The authors would state that although the simulations were carried out with changed values of Δg, the newly obtained results do not deviate largely from the original publication. The trends for phase fractions with regard to the simulation time steps, normalized temperatures and normalized shear stresses are qualitatively the same as the results in the original publication, and thus the conclusions in the original publication were not affected. The authors would like to apologise for any inconvenience caused.