Self-Adaptive Gate Control for Efficient Escape From Local Minimum Energy on Invertible Logic

Invertible logic can realize bidirectional operations of a function represented by Hamiltonian (energy) with noise and be applied for integer factorization training neural networks. However, the computation error is not negligible due to becoming trapped at local minimum energy when large. This paper introduces self-adaptive gate control (SAGC) high convergence rates global on invertible logic. From our analysis based simulating small-scale circuits, it supposed that stuck caused invalid states gates. The proposed SAGC autonomously detects gates checking truth tables adds large them efficient escape from energy. As typical example logic, adders, multipliers multiplexers are designed evaluated. simulation results show probabilities few times higher than those conventional method equally all