diff --git a/vllm/v1/attention/backends/mla/rocm_aiter_mla.py b/vllm/v1/attention/backends/mla/rocm_aiter_mla.py
index f46010d757af..3abb185c5b8f 100644
--- a/vllm/v1/attention/backends/mla/rocm_aiter_mla.py
+++ b/vllm/v1/attention/backends/mla/rocm_aiter_mla.py
@@ -98,17 +98,17 @@ def _get_paged_kv_tensors(
             paged_kv_last_page_len,
         )
 
-    def _build_decode(self, block_table: torch.Tensor,
+    def _build_decode(self, block_table_tensor: torch.Tensor,
                       seq_lens: torch.Tensor) -> AiterMLADecodeMetadata:
 
         (
             paged_kv_indices,
             paged_kv_indptr,
             paged_last_page_len,
-        ) = self._get_paged_kv_tensors(block_table, seq_lens)
+        ) = self._get_paged_kv_tensors(block_table_tensor, seq_lens)
 
         attn_metadata = AiterMLADecodeMetadata(
-            block_table=block_table,
+            block_table=block_table_tensor,
             seq_lens=seq_lens,
             paged_kv_indptr=paged_kv_indptr,
             paged_kv_indices=paged_kv_indices,