Add MoE scale ratio output

tests/test_multilayer.py

@@ -146,7 +146,8 @@ def main():
         cos = F.cosine_similarity(ref_hidden.flatten().unsqueeze(0), run_hidden.flatten().unsqueeze(0)).item()
         has_nan = torch.isnan(run_hidden).any().item()
         has_inf = torch.isinf(run_hidden).any().item()
-        print(f"Layer {layer}: MoE_cosine={cos_moe:.6f} Residual_cosine={cos:.6f} ref_amax={ref_hidden.amax().item():.4f} run_amax={run_hidden.amax().item():.4f} NaN={has_nan} Inf={has_inf}")
+        moe_scale = run_out.abs().mean().item() / max(ref_out.abs().mean().item(), 1e-8)
+        print(f"Layer {layer}: MoE_cosine={cos_moe:.6f} MoE_scale={moe_scale:.4f} ref_moe_amax={ref_out.amax().item():.4f} run_moe_amax={run_out.amax().item():.4f} NaN={has_nan}")
         
         if has_nan:
             print(f"  ❌ NaN detected after layer {layer}! Stopping.")