diff --git a/onnxruntime/core/mlas/lib/dwconv.cpp b/onnxruntime/core/mlas/lib/dwconv.cpp
index 19951a629f847..87effeac5fd9f 100644
--- a/onnxruntime/core/mlas/lib/dwconv.cpp
+++ b/onnxruntime/core/mlas/lib/dwconv.cpp
@@ -14,7 +14,6 @@ Module Name:
 
 --*/
 
-
 #include "fp16_common.h"
 
 #ifdef MLAS_F16VEC_INTRINSICS_SUPPORTED
@@ -30,7 +29,7 @@ MlasConvDepthwiseKernel(
     size_t OutputCount,
     size_t KernelSize,
     MLAS_HALF_GEMM_POSTPROCESSOR* PostProc
-    )
+)
 {
     while (OutputCount > 0) {
         size_t ChannelOffset = 0;
@@ -87,8 +86,7 @@ MlasConvDepthwiseKernel(
             Output += c;
         }
         if (PostProc) {
-            PostProc->Process(reinterpret_cast<MLAS_FP16*>(Output - Channels), 0, 0, 1, Channels,
-                              Channels);
+            PostProc->Process(reinterpret_cast<MLAS_FP16*>(Output - Channels), 0, 0, 1, Channels, Channels);
         }
         Input += KernelSize;
         OutputCount -= 1;
@@ -108,7 +106,7 @@ MlasConvDepthwiseKernel(
     size_t OutputCount,
     size_t KernelSize,
     MLAS_HALF_GEMM_POSTPROCESSOR* PostProc
-    )
+)
 {
     while (OutputCount > 0) {
         for (size_t ChannelOffset = 0; ChannelOffset < Channels; ChannelOffset++) {
@@ -122,16 +120,14 @@ MlasConvDepthwiseKernel(
             *Output++ = MLAS_Float2Half(Accumulator);
         }
         if (PostProc) {
-            PostProc->Process(reinterpret_cast<MLAS_FP16*>(Output - Channels), 0, 0, 1, Channels,
-                              Channels);
+            PostProc->Process(reinterpret_cast<MLAS_FP16*>(Output - Channels), 0, 0, 1, Channels, Channels);
         }
         Input += KernelSize;
         OutputCount -= 1;
     }
 }
 
-#endif // MLAS_F16VEC_INTRINSICS_SUPPORTED
-
+#endif  // MLAS_F16VEC_INTRINSICS_SUPPORTED
 
 void
 MLASCALL
@@ -144,7 +140,7 @@ MlasConvDepthwise(
     size_t OutputCount,
     size_t KernelSize,
     MLAS_HALF_GEMM_POSTPROCESSOR* PostProc
-    )
+)
 {
     MlasConvDepthwiseKernel(
         reinterpret_cast<const _mlas_fp16_* const*>(Input),
@@ -154,5 +150,6 @@ MlasConvDepthwise(
         Channels,
         OutputCount,
         KernelSize,
-        PostProc);
+        PostProc
+    );
 }
diff --git a/onnxruntime/test/providers/cpu/nn/conv_fp16_test.cc b/onnxruntime/test/providers/cpu/nn/conv_fp16_test.cc
index 5a2660c05fe35..f723bd10ec581 100644
--- a/onnxruntime/test/providers/cpu/nn/conv_fp16_test.cc
+++ b/onnxruntime/test/providers/cpu/nn/conv_fp16_test.cc
@@ -732,8 +732,7 @@ TEST(ConvFp16Test, Depthwise2D_Bias) {
 
       MLFloat16(9.0f), MLFloat16(10.0f), MLFloat16(11.0f),
       MLFloat16(12.0f), MLFloat16(13.0f), MLFloat16(14.0f),
-      MLFloat16(15.0f), MLFloat16(16.0f), MLFloat16(17.0f)
-  };
+      MLFloat16(15.0f), MLFloat16(16.0f), MLFloat16(17.0f)};
   vector<int64_t> X_shape = {1, 2, 3, 3};
   vector<MLFloat16> W = {MLFloat16(1.0f), MLFloat16(2.0f)};
   vector<int64_t> W_shape = {2, 1, 1, 1};
@@ -747,8 +746,7 @@ TEST(ConvFp16Test, Depthwise2D_Bias) {
 
       MLFloat16(17.0f), MLFloat16(19.0f), MLFloat16(21.0f),
       MLFloat16(23.0f), MLFloat16(25.0f), MLFloat16(27.0f),
-      MLFloat16(29.0f), MLFloat16(31.0f), MLFloat16(33.0f)
-  };
+      MLFloat16(29.0f), MLFloat16(31.0f), MLFloat16(33.0f)};
 
   TestConvFp16Op(attrs, {X, W, B}, {X_shape, W_shape, B_shape}, expected_vals, Y_shape);
   TestConvFp16Op(attrs, {X, W, B}, {X_shape, W_shape, B_shape}, expected_vals, Y_shape, true);
@@ -874,16 +872,36 @@ TEST(ConvFp16Test, Depthwise2D_Bias_Complex) {
   };
   vector<int64_t> W_shape = {13, 1, 2, 2};
   vector<MLFloat16> B = {
-      MLFloat16(1.0f), MLFloat16(2.0f), MLFloat16(3.0f), MLFloat16(4.0f), MLFloat16(5.0f), MLFloat16(6.0f),
-      MLFloat16(7.0f), MLFloat16(8.0f), MLFloat16(9.0f), MLFloat16(10.0f), MLFloat16(11.0f), MLFloat16(12.0f),
+      MLFloat16(1.0f),
+      MLFloat16(2.0f),
+      MLFloat16(3.0f),
+      MLFloat16(4.0f),
+      MLFloat16(5.0f),
+      MLFloat16(6.0f),
+      MLFloat16(7.0f),
+      MLFloat16(8.0f),
+      MLFloat16(9.0f),
+      MLFloat16(10.0f),
+      MLFloat16(11.0f),
+      MLFloat16(12.0f),
       MLFloat16(13.0f),
   };
   vector<int64_t> B_shape = {13};
   vector<int64_t> Y_shape = {1, 13, 1, 1};
   auto expected_vals = {
-      MLFloat16(15.0f), MLFloat16(128.0f), MLFloat16(369.0f), MLFloat16(738.0f), MLFloat16(1235.0f),
-      MLFloat16(1860.0f), MLFloat16(2613.0f), MLFloat16(3494.0f), MLFloat16(4503.0f), MLFloat16(5640.0f),
-      MLFloat16(6905.0f), MLFloat16(8298.0f), MLFloat16(9819.0f),
+      MLFloat16(15.0f),  // 0.0*0.0 + 1.0*1.0 + 2.0*2.0 + 3.0*3.0 + 1.0
+      MLFloat16(128.0f),
+      MLFloat16(369.0f),
+      MLFloat16(738.0f),
+      MLFloat16(1235.0f),
+      MLFloat16(1860.0f),
+      MLFloat16(2613.0f),  // 24.0*24.0 + 25.0*25.0 + 26.0*26.0 + 27.0*27.0 + 7.0
+      MLFloat16(3494.0f),
+      MLFloat16(4503.0f),
+      MLFloat16(5640.0f),
+      MLFloat16(6905.0f),
+      MLFloat16(8298.0f),
+      MLFloat16(9819.0f),  // 48.0*48.0 + 49.0*49.0 + 50.0*50.0 + 51.0*51.0 + 13.0
   };
 
   TestConvFp16Op(attrs, {X, W, B}, {X_shape, W_shape, B_shape}, expected_vals, Y_shape);
@@ -1252,4 +1270,4 @@ TEST(ConvFp16Test, SharedPrepackedWeights) {
 }  // namespace test
 }  // namespace onnxruntime
 
-#endif  // MLAS_F16VEC_INTRINSICS_SUPPORTED
\ No newline at end of file
+#endif  // MLAS_F16VEC_INTRINSICS_SUPPORTED
diff --git a/onnxruntime/test/providers/cpu/nn/conv_op_test.cc b/onnxruntime/test/providers/cpu/nn/conv_op_test.cc
index 3905019cf57d6..8cf8c027bdde2 100644
--- a/onnxruntime/test/providers/cpu/nn/conv_op_test.cc
+++ b/onnxruntime/test/providers/cpu/nn/conv_op_test.cc
@@ -796,7 +796,19 @@ TEST(ConvTest, Depthwise2D_Bias_Complex) {
   vector<int64_t> B_shape = {13};
   vector<int64_t> Y_shape = {1, 13, 1, 1};
   auto expected_vals = {
-      15.0f, 128.0f, 369.0f, 738.0f, 1235.0f, 1860.0f, 2613.0f, 3494.0f, 4503.0f, 5640.0f, 6905.0f, 8298.0f, 9819.0f
+      15.0f,  // 0.0*0.0 + 1.0*1.0 + 2.0*2.0 + 3.0*3.0 + 1.0
+      128.0f,
+      369.0f,
+      738.0f,
+      1235.0f,
+      1860.0f,
+      2613.0f,  // 24.0*24.0 + 25.0*25.0 + 26.0*26.0 + 27.0*27.0 + 7.0
+      3494.0f,
+      4503.0f,
+      5640.0f,
+      6905.0f,
+      8298.0f,
+      9819.0f,  // 48.0*48.0 + 49.0*49.0 + 50.0*50.0 + 51.0*51.0 + 13.0
   };
 
   TestConvOp(attrs, {X, W, B}, {X_shape, W_shape, B_shape}, expected_vals, Y_shape);