• In the mixed-adaptive model, multiplexing gains the thresholds forward and backward transitions against a monolayer problem. • The hysteresis in a multiplex network is observed at a larger population of competitive units against a single-layer model. • The multiplex configuration is more robust against self-consistent amplification of pre-bifurcation bursts of local coherence. In dynamical networks, the presence of adaptation establishing the relationship between the coherence of local populations and unit’s effective coupling provides the explosive transition — an abrupt transition from incoherence to coherence and vice versa through the hysteresis loop. Explosive transition is even possible under the coexistence of two opposite types of adaptation – interdependence and competition, wherein growing the competitive population dramatically narrows the area of hysteresis. Here, we demonstrate that considering a mixed adaptive model from a multilayer perspective expands the hysteresis region and shifts both forward and backward transition boundaries to the higher values of coupling strength as compared with a monolayer case. We show that this is due to greater robustness of the multilayer network against the intralayer topology and lower sensitivity to the amplification of the pre-bifurcation noise, i.e., spurious fluctuations of local coherence, in the vicinity of a tipping point as opposed to a single-layer network.