Tectonic faults are characterized by depth-dependent frictional, hydraulic and structural properties, with seismic and aseismic processes interacting near the base of the seismogenic zone. While many fault segments feature abundance of microseismicity at depth, some major strike-slip faults known to have had large earthquakes are silent in the interseismic period, posing a long-standing enigma. To understand such phenomena and the relation between co-, post- and inter-seismic behavior of seismogenic faults, we study earthquake sequences and aseismic slip in 3D fault models that incorporate laboratory-derived rate-and-state friction laws and enhanced dynamic weakening mechanisms. Based on our physical models as well as observations of recent major events, we suggest that absence of concentrated microseismicity at depth is an indicator that previous large earthquakes penetrated deeper below the conventionally defined seismogenic zone. For the Carrizo segment of the San Andreas Fault (SAF) in Southern California, which hosted the 1857 M_w 7.9 Fort Tejon earthquake, we infer a penetration distance of at least 3-5 km. The propensity of such deeper penetration has important implications for deep fault rheology and earthquake scaling relation. In contrast, fault segments such as the Parkfield on the SAF and those on the San Jacinto Fault, which have concentrated microseismicity in the interseismic period at the transitional depths, are likely to produce major events with ruptures limited within the seismogenic zone and hence smaller sizes. In both cases, the spatio-temporal complexity of small earthquakes in the post- and inter-seismic periods, if any, would reflect and constrain the state of fault coupling within the deeper locked-creeping transition.