This article reports the direct observation of conducting filament growth in nanoscale resistive memories using transmission electron microscopy (TEM) and structural analysis. The study reveals that the growth of conducting filaments is primarily driven by cation transport in the dielectric film. Two distinct growth modes were observed in materials with different microstructures, with the narrowest region of the filament located near the dielectric/inert-electrode interface. This finding suggests that the dielectric/inert electrode interface is critical for device optimization. Nanoscale resistive switching devices, often termed memristors, have been studied for applications in memory, logic, and neuromorphic computing. The resistive switching effect is typically attributed to the formation of conducting filaments, but the nature and growth dynamics of these filaments remain unclear. The study provides unambiguous evidence that conducting filaments composed of elemental metals (and sometimes nanoparticles) are formed during resistive switching. The growth of these filaments is influenced by cation mobility, and the study identifies two different growth modes, including one where the filament starts from the active electrode and grows towards the inert electrode, which contradicts previous theories. The study also examines the erasing mechanism, finding that the dissolution of conducting filaments occurs near the inert electrode interface, which is unexpected in previous hypotheses. The results were supported by in-situ TEM studies on memory devices with vertical configurations. The findings highlight the importance of cation transport in resistive switching and suggest that the dielectric/inert electrode interface is critical for device operation. The study also explores the role of cation transport in dielectric-based resistive memories, showing that the growth of conducting filaments can be limited by cation supply. The results indicate that the filament growth dynamics are significantly influenced by the choice of dielectric material. The study further demonstrates that the filament growth can be controlled by adjusting the programming current, with higher currents leading to larger filament sizes. The study also investigates the electric field and temperature dependence of the wait time before resistance switching, finding that the filament growth is facilitated by the electric field. The results support the hypothesis that the cation transport process is the rate-limiting factor in these devices. The study provides important insights into the switching mechanism in filament-based resistive memories and highlights the need for continued material and device optimizations.This article reports the direct observation of conducting filament growth in nanoscale resistive memories using transmission electron microscopy (TEM) and structural analysis. The study reveals that the growth of conducting filaments is primarily driven by cation transport in the dielectric film. Two distinct growth modes were observed in materials with different microstructures, with the narrowest region of the filament located near the dielectric/inert-electrode interface. This finding suggests that the dielectric/inert electrode interface is critical for device optimization. Nanoscale resistive switching devices, often termed memristors, have been studied for applications in memory, logic, and neuromorphic computing. The resistive switching effect is typically attributed to the formation of conducting filaments, but the nature and growth dynamics of these filaments remain unclear. The study provides unambiguous evidence that conducting filaments composed of elemental metals (and sometimes nanoparticles) are formed during resistive switching. The growth of these filaments is influenced by cation mobility, and the study identifies two different growth modes, including one where the filament starts from the active electrode and grows towards the inert electrode, which contradicts previous theories. The study also examines the erasing mechanism, finding that the dissolution of conducting filaments occurs near the inert electrode interface, which is unexpected in previous hypotheses. The results were supported by in-situ TEM studies on memory devices with vertical configurations. The findings highlight the importance of cation transport in resistive switching and suggest that the dielectric/inert electrode interface is critical for device operation. The study also explores the role of cation transport in dielectric-based resistive memories, showing that the growth of conducting filaments can be limited by cation supply. The results indicate that the filament growth dynamics are significantly influenced by the choice of dielectric material. The study further demonstrates that the filament growth can be controlled by adjusting the programming current, with higher currents leading to larger filament sizes. The study also investigates the electric field and temperature dependence of the wait time before resistance switching, finding that the filament growth is facilitated by the electric field. The results support the hypothesis that the cation transport process is the rate-limiting factor in these devices. The study provides important insights into the switching mechanism in filament-based resistive memories and highlights the need for continued material and device optimizations.