Comparison of the star-forming main sequence at different redshifts when adopting an evolving SMF in input. Upper panel: Main sequence predicted from decode for |$z = 0, 0.6, 1, 1.5 \,\, {\rm and} \,\, 2$| (solid lines). The dashed lines show the extreme case where mergers are absent. Central panel: Comparison between the main sequence at |$z=1$| from decode and those from Leja et al. (2022), Popesso et al. (2023), Koprowski et al. (2024), and the TNG simulation, as labelled. Lower panel: Same as central panel but for |$z=2$|⁠.