diff --git a/llmc/rmsnorm.cuh b/llmc/rmsnorm.cuh
index 8f20e98..5be05bb 100644
--- a/llmc/rmsnorm.cuh
+++ b/llmc/rmsnorm.cuh
@@ -90,7 +90,7 @@ __global__ void fused_residual_rmsnorm_forward_kernel5(floatX* residual, floatX*
     __syncthreads();
 
     int idx = blockIdx.x * blockDim.y + threadIdx.y;
-    if(idx > N) return;
+    if(idx >= N) return;
 
     // adjust pointers to current token
     residual += C * idx;