The Galicia magma-poor margin is considered an archetypal location to explore detachment faulting in a passive rift margin setting. A new 3-D seismic reflection volume acquired over the Galicia continent-ocean transition zone provides an unprecedented view of the prominent S-reflector detachment fault from the Deep Galicia Margin. This new dataset shows that the interface of the S-reflector detachment fault has heterogenous seismic reflection amplitudes. The seismic reflection amplitudes are analyzed and discussed relative to fault surface morphology and variations in the fault rock layer above and upper mantle layer below the S-reflector. The fault surface map of the S-reflector shows coherent corrugations parallel to the expected paleo-extension directions with an average azimuth of 107o. Locally, the corrugations are continuous beneath multiple fault blocks, spanning crustal fault intersections with the S-reflector, which suggest that during the final stages of the detachment did not demonstrate sequential crustal faulting. A second somewhat discontinuous layer above the detachment fault, here named the S-interval, interpreted to be as zone of fault rock above the S-reflector. Overall, the S-interval increases in thickness by tens of meters to the northwest, in the direction of hanging wall transport. There are localized thick accumulations of fault rock near overlying fault intersections, suggesting either non-uniform fault rock production, or redistribution of fault rock during slip. An artificial neural network has been utilized to estimate the P-wave velocities of the upper mantle, 100 ms (~400 m) below the S-reflector detachment. This approach estimates seismic velocity along a target interface, based on geological and geophysical inputs from both the 3-D seismic reflection survey and and additional 2-D wide-angle seismic profile. This non-linear regression technique results in estimated velocities 100 ms above and below the targeted interface, the S-reflector. An ensemble of similarly trained neural networks is used to estimate previously unknown target velocities where the S-reflector is the cust-mantle boundary within the 3-D dataset, where the S-reflector is the crust-mantle boundary. These networks reached minimum achievable error of 2% on the test data, and the low standard deviation (<300 m/s) between different networks demonstrate that the input features were sufficient to capture variations in the velocity above and below the targeted S-reflector. The upper mantle seismic velocities are used to estimate distribution and the degree of serpentinization in the upper mantle and its distribution beneath the S-reflector detachment fault. The degree of serpentinization is heterogeneous, varying between 0 and 90% across the domain. The distribution pattern shows relationships with both the current intersections of crust-cutting faults soles and to paleo-temperatures estimated for ~112 Ma. High degrees of serpentinization are only observed in areas where overburden thickness at the end of rifting was sufficient to provide optimal serpentinization temperatures. The alignment of crust-cutting fault intersections and serpentinized areas suggest the crustal faults were water conduits that hydrated the mantle peridotite. The correspondence between the fault intersections and present-day serpentinite anomalies indicates that the most of the serpentinization postdated slip on the S-reflector detachment. Volumetric expansion caused by the serpentinization process may also lead to undulations on the fault detachment fault surface. High localized reliefs (0.9 km and 1.4 km) on the S-reflector coincide with highly serpentinized (80-90%) areas in the 3-D seismic reflection survey. These observations have important implications for understanding how detachment faults form and evolve during and after rifting. 3-D seismic reflection imaging and amplitude analysis have enabled unique insights into fault slip history, fault rock production and redistribution, and causal mechanisms behind mantle serpentinization in a passive rift margin environment.