A Leaky Integrate-and-Fire (LIF) neuron is a simple mathematical model used in neuroscience and computational neuroscience to describe the behavior of individual neurons. It provides a simplified yet effective way to simulate the behavior of real neurons. In the neuron, there are two key elements: the current and the threshold. If the current surpasses the threshold, a spike is emitted, otherwise, the spike is 0 (resting). In this spiking LSTM implementation, the neuron's threshold is adaptive, and will increase when the threshold is passed. When the neuron fails to reach the threshold, it will slowy decay back to the initial threshold.
After applying the reset, the variables should be initialized, and a current can be applied. The testbench will record the current, threshold, and spiking behavior of the neuron. When a higher current is applied, you can see how the threshold increases, and vice versa when a spike is not emitted.
| # | Input | Output | Bidirectional |
|---|---|---|---|
| 0 | clk Clock input | uo_out Spike output | uio_in Unused |
| 1 | rst_n Reset signal | uio_out Threshold | |
| 2 | ui_in Voltage current | uio_oe Unused | |
| 3 | ena Unused | ||
| 4 | |||
| 5 | |||
| 6 | |||
| 7 |