Abstract
Stretchable sensor arrays are essential for bridging the physical environment with intelligent tactile perception in wearable electronics and soft robotics. However, achieving high sensitivity across a wide dynamic range remains a persistent challenge due to the inherent trade-offs in conductive network stability. Here, we report an architected sensor array featuring tunable sensitivity, enabled by a mixed-dimensional network comprising carbon nanofibers (CNFs) and carbon black (CB). Diverging from single-filler systems characterized by abrupt conductivity transitions, this synergistic design employs CNFs as a structural skeleton and CB particles as interconnected tunneling bridges. Such geometric coupling effectively lowers the percolation threshold and establishes a gradual transition zone, which preserves mechanical compliance while ensuring stable electromechanical transduction. To mitigate crosstalk in large-scale matrices, a hardware-software integrated strategy is implemented, combining multi-channel scanning with an optimized equivalent circuit model. This approach facilitates algorithmic signal decoupling for high-fidelity pressure mapping and precise reconstruction of handwritten trajectories. This work provides a scalable strategy for engineering high-performance soft electronics with enhanced tactile intelligence.