Figure 4 | Scientific Reports

Figure 4

From: Machine learning-powered compact modeling of stochastic electronic devices using mixture density networks

Figure 4

Switching probabilities at varying bias currents and transient simulation results. The predicted vs experimental switching probability for gate current (\(I_G\)) between 0 \(\mu A\) and 3 \(\mu A\) at a bias current (\(I_B\)) of (a) 14 \(\mu A\), (b) 16.5 \(\mu A\), (c) 23.5 \(\mu A\), (d) 28 \(\mu A\), and (e) 33 \(\mu A\). The increase in \(I_B\) leads to a reduction in the gate switching current. Transient dynamics of (f) bias current and (g) gate current which cause switching of the channel after exceeding the corresponding thresholds. (h) Switching is shown and validated with the experiment by the time dynamics of the load voltage (\(V_L\)) at mean (\(\mu\)) and plus or minus 2 standard deviations (\(\sigma\)). (i) A zoomed-in view of \(V_L\) shown in (h).

Back to article page