The provided code is a simulation of a computational model of neuronal networks with a focus on understanding orientation selectivity in inhibition-dominated networks of spiking neurons. This model is relevant to understanding how cortical networks process visual information, particularly within the primary visual cortex (V1) of the brain. ### Biological Basis #### 1. **Neuron Types and Dynamics** - **Leaky Integrate-and-Fire (LIF) vs. Perfect Integrate-and-Fire (PIF) Neurons:** The model supports simulation using either LIF or PIF neurons. LIF neurons are more biologically realistic as they feature a membrane potential that exponentially decays over time, mimicking the leaky nature of real neurons. PIF neurons do not account for this leakage, signifying an idealized neuron type. The parameter `tauMem` in the model sets the membrane time constant, specifying how quickly the membrane potential decays, which affects neuronal excitability and response dynamics. #### 2. **Network Structure** - **Excitatory and Inhibitory Populations:** The model considers both excitatory and inhibitory connections, with parameters such as `eps_exc` (excitatory connection probability) and `eps_inh` (inhibitory connection probability) defining the network's connectivity structure. This reflects the balanced nature of cortical networks where excitatory and inhibitory neurons interact intricately to produce stable network activity. #### 3. **Delays and Dynamics** - **Synaptic Delays:** The model incorporates synaptic delays, either fixed or randomized, with parameters being set based on biological variations in synaptic transmission times. Synaptic delays are crucial for temporal dynamics and information processing in neural circuits. #### 4. **Stimuli and Responses** - **Orientation Selectivity:** The primary focus of the model is on orientation selectivity, the ability of neurons in V1 to respond preferentially to edges at specific orientations in the visual field. The model simulates responses to stimuli at different orientations (`stim_range`) and analyzes the network's ability to produce selective responses to these stimuli. - **Contrast Sensitivity:** The model also considers varying levels of stimulus contrast (`contrast` parameter), examining how network activity and orientation selectivity change with different strengths of input signals. #### 5. **Spontaneous and Evoked Activity** - **Spontaneous Activity:** The simulation first examines the network's spontaneous activity, without orientation-specific stimuli, to establish a baseline measure of the network's dynamics in a resting state. - **Evoked Responses:** The network's response to specific stimulus orientations is then analyzed, reflecting evoked neuronal activity as seen in real visual processing where external stimuli modulate neural responses. This computational model captures essential features of the biological visual system, including excitatory-inhibitory balance, synaptic connectivity, temporal dynamics, and input-driven modulation of activity, offering insights into the underlying mechanisms of orientation selectivity in the visual cortex.