last changes

mathema/actors/neuron.py

@@ -12,6 +12,24 @@ def tanh(x): return math.tanh(x)
 
 
 class Neuron(Actor):
+    """
+    Neuron actor implementing a weighted-sum neuron with an activation function
+    and optional recurrent inputs.
+
+    The Neuron receives input vectors from upstream neurons, accumulates them
+    according to its weight configuration, applies an activation function,
+    and forwards the resulting output to downstream actors.
+
+    It supports:
+        - feed-forward and recurrent connections
+        - bias handling
+        - asynchronous message-based execution
+        - weight backup, restoration, and stochastic perturbation (mutation)
+        - cycle-based updates for recurrent networks
+
+    This actor is designed to be used inside a cortex/agent actor network,
+    where synchronization and evolution are coordinated externally.
+    """
     def __init__(self, nid, cx_pid, af_name, input_idps, output_pids, bias: Optional[float] = None):
         super().__init__(f"Neuron-{nid}")
         self.nid = nid