Update node.py

node.py

@@ -11,17 +11,23 @@ class CognitiveNode(nn.Module):
         self.activation = 0.0
         
         # Dynamic input weights with Hebbian plasticity
-        self.weights = nn.Parameter(torch.randn(input_size) * 0.1)
+        self.weights = nn.Parameter(torch.randn(1))  # Changed from input_size to 1
         self.bias = nn.Parameter(torch.zeros(1))
         
-        # Memory system
-        self.memory = CognitiveMemory(context_size=input_size)
+        # Memory system - adjusted context size
+        self.memory = CognitiveMemory(context_size=1)  # Changed from input_size to 1
         
         # Neurotransmitter levels
         self.dopamine = nn.Parameter(torch.tensor(0.5))
         self.serotonin = nn.Parameter(torch.tensor(0.5))
         
+        # Store recent activations
+        self.recent_activations = {}
+        
     def forward(self, inputs: torch.Tensor) -> torch.Tensor:
+        # Ensure inputs is a single value tensor
+        inputs = inputs.reshape(1)
+        
         # Memory influence
         mem_context = self.memory.retrieve(inputs)
         
@@ -29,12 +35,17 @@ class CognitiveNode(nn.Module):
         combined = inputs * 0.7 + mem_context * 0.3
         
         # Adaptive activation with neurotransmitter modulation
-        base_activation = torch.tanh(combined @ self.weights + self.bias)
+        base_activation = torch.tanh(combined * self.weights + self.bias)
         modulated = base_activation * (1 + self.dopamine - self.serotonin)
         
         # Update memory
         self.memory.add_memory(inputs, modulated.item())
         
+        # Store recent activation
+        self.recent_activations[len(self.recent_activations)] = modulated.item()
+        if len(self.recent_activations) > 100:
+            self.recent_activations.pop(min(self.recent_activations.keys()))
+        
         return modulated
     
     def update_plasticity(self, reward: float):