V1 Six-layer Column¶
A laminar model inspired by primate V1 with six layers and depth-specific readouts.
Configuration¶
import jaxfne as jtfne
cfg = (
jtfne.configuration()
.network(
n=600,
layers=["L1", "L2/3", "L4", "L5", "L6"],
cell_types={"E": 0.8, "PV": 0.1, "SST": 0.07, "VIP": 0.03},
connectivity="layer_structured"
)
.emitter(family="izhikevich", preset="cortical_eig")
.field(
domain="laminar_column",
conductivity="proxy",
depths=[0.0, 0.15, 0.3, 0.5, 0.7, 1.0]
)
.probe(
name="v1_column",
n_contacts=6,
modes=["spikes", "V_m", "LFP", "CSD"]
)
)
model = jtfne.construct(cfg)
Multimodal readouts¶
signals = model.simulate(...)
# Layer-specific rates
readouts = model.compute_readout(signals, [
jtfne.readout_spec("L4_rate", "spike_rate_hz"),
jtfne.readout_spec("LFP_L4", "lfp_abs_mean"),
jtfne.readout_spec("CSD_L4", "csd_abs_mean"),
])
Laminar features¶
- L4 receives thalamocortical input (configurable)
- L2/3, L5, L6 have inter-laminar projections
- LFP-proxy reflects population summed current
- CSD-proxy reflects current source densities per layer
Next step¶
Progress to V1-PFC dual column for multi-areal networks.