diff --git a/vllm/core/block/cpu_offloading_block_allocator.py b/vllm/core/block/cpu_offloading_block_allocator.py
index 41a6e94c4e58c..93ebbd479ff00 100644
--- a/vllm/core/block/cpu_offloading_block_allocator.py
+++ b/vllm/core/block/cpu_offloading_block_allocator.py
@@ -109,10 +109,13 @@ def __init__(self, cpu_block_allocator: PrefixCachingBlockAllocator,
         """
         GPU block should only be in one of the following three status:
           uncached: allocated blocks that didn't hit any cache
-          cached: allocated blocks that hit the cache, either in GPU or in CPU
+          cached: allocated blocks that are cached, either in GPU or in CPU
           free: the blocks are not allocated by block allocator
-        block allocator will automatically track free blocks, and we don't need
-        to specially handle cached blocks. So we only track uncached blocks
+        This implementation aims to transform uncacherd blocks to cached blocks
+        by performing GPU to CPU copy when calling `get_and_reset_swaps`
+        
+        As block allocator will automatically track free blocks, and we don't 
+        need to specially handle cached blocks. So we only track uncached blocks
         """
         self._uncached_blocks: Deque[Block] = deque()
         """