“…Direct characterization of the internal structure of brittle fault zones in the subsurface is limited to sparse borehole data, while seismic and other geophysical methods (e.g., tomography based on different observables) can provide high‐quality 2‐D or 3‐D characterizations but still affected by noise and resolution limits (e.g., Amoroso et al, 2017; Ben‐Zion, 1998; Cochran et al, 2009; Hale, 2013; Iacopini et al, 2016; Pischiutta et al, 2017; Vidale & Li, 2003). In the case of seismogenic fault zones, modern techniques of aftershock relocation (i.e., double difference method and earthquake detection algorithms with deep neural networks) allowed seismologists to illuminate with unprecedented internal spatial resolution (~10 m of relocation precision for correlated events) the three‐dimensional fault networks activated during seismic sequences (Chiaraluce et al, 2011; Ross et al, 2020; Valoroso et al, 2013).…”