diff --git a/.github/workflows/auto-merge.yml b/.github/workflows/auto-merge.yml
index 7840644e..a3dc6dc3 100644
--- a/.github/workflows/auto-merge.yml
+++ b/.github/workflows/auto-merge.yml
@@ -8,6 +8,10 @@ on:
     branches:
       - main  # 仅在针对main分支的PR上运行
 
+permissions:
+  contents: write
+  pull-requests: write
+
 jobs:
   auto-merge:
     runs-on: ubuntu-latest
@@ -24,4 +28,4 @@ jobs:
           MERGE_RETRIES: "6"
           MERGE_RETRY_SLEEP: "10000"
           UPDATE_LABELS: "auto-merge"
-          UPDATE_METHOD: "merge" 
\ No newline at end of file
+          UPDATE_METHOD: "rebase" 
\ No newline at end of file
diff --git a/.github/workflows/excuter-cuda-linux.yml b/.github/workflows/excuter-cuda-linux.yml
index d523bca7..b65169c4 100644
--- a/.github/workflows/excuter-cuda-linux.yml
+++ b/.github/workflows/excuter-cuda-linux.yml
@@ -57,7 +57,7 @@ jobs:
             cd /workspace && \
             
             # 构建 common 库
-            cd excuter/common && \
+            cd excuter/cpp-common && \
             mkdir -p build && cd build && \
             cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER_LAUNCHER=ccache -GNinja .. && \
             ninja && \
diff --git a/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu b/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu
index e399cf66..aaeed355 100644
--- a/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu
+++ b/excuter/op-mem-cuda/src/deepx/tensorfunc/init_miaobyte.cu
@@ -50,7 +50,7 @@ namespace deepx::tensorfunc
         int idx = blockIdx.x * blockDim.x + threadIdx.x;
         if (idx < size)
         {
-            data[idx] = start + step * static_cast<T>(idx);
+            data[idx] = start + step * static_cast<T>(static_cast<float>(idx));
         }
     }
     template __global__ void kernel_arange<double>(double *data, const int size, const double start, const double step);