diff --git a/vllm_metax/models/deepseek_v2.py b/vllm_metax/models/deepseek_v2.py
index 634819d16..ea058b122 100644
--- a/vllm_metax/models/deepseek_v2.py
+++ b/vllm_metax/models/deepseek_v2.py
@@ -632,7 +632,7 @@ def sparse_attn_indexer(
             decode_metadata.seq_lens,
             decode_metadata.block_table,
             decode_metadata.schedule_metadata,
-            max_context_len=max_model_len,
+            max_model_len,
         )
         # padded query len
         current_device = padded_q_bf16_decode_tokens.device
diff --git a/vllm_metax/patch/__init__.py b/vllm_metax/patch/__init__.py
index 27a76f597..6a5564c75 100644
--- a/vllm_metax/patch/__init__.py
+++ b/vllm_metax/patch/__init__.py
@@ -6,3 +6,4 @@
 from . import device_allocator
 from . import model_executor
 from . import oot
+from . import sample
\ No newline at end of file
diff --git a/vllm_metax/patch/sample/__init__.py b/vllm_metax/patch/sample/__init__.py
new file mode 100644
index 000000000..bc96e2874
--- /dev/null
+++ b/vllm_metax/patch/sample/__init__.py
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier: Apache-2.0
+
+from . import rejection_sampler
diff --git a/vllm_metax/patch/sample/rejection_sampler.py b/vllm_metax/patch/sample/rejection_sampler.py
new file mode 100644
index 000000000..9355453ee
--- /dev/null
+++ b/vllm_metax/patch/sample/rejection_sampler.py
@@ -0,0 +1,86 @@
+# SPDX-License-Identifier: Apache-2.0
+from vllm.triton_utils import tl, triton
+
+import vllm.v1.sample.rejection_sampler
+
+# SPDX-License-Identifier: Apache-2.0
+
+
+@triton.jit
+def sample_recovered_tokens_kernel(
+    output_token_ids_ptr,  # [num_tokens]
+    cu_num_draft_tokens_ptr,  # [batch_size]
+    draft_token_ids_ptr,  # [num_tokens]
+    draft_probs_ptr,  # [num_tokens, vocab_size] or None
+    target_probs_ptr,  # [num_tokens, vocab_size]
+    q_ptr,  # [batch_size, vocab_size]
+    vocab_size,
+    PADDED_VOCAB_SIZE: tl.constexpr,
+    NO_DRAFT_PROBS: tl.constexpr,
+    BLOCK_SIZE: tl.constexpr = 1024,
+):
+    req_idx = tl.program_id(0)
+    if req_idx == 0:
+        start_idx = 0
+    else:
+        start_idx = tl.load(cu_num_draft_tokens_ptr + req_idx - 1)
+    end_idx = tl.load(cu_num_draft_tokens_ptr + req_idx)
+    num_draft_tokens = end_idx - start_idx
+
+    # Early exit for out-of-range positions.
+    pos = tl.program_id(1)
+    if pos >= num_draft_tokens:
+        return
+
+    max_prob = -float('inf')
+    best_token_id = 0
+
+    for block_start in range(0, PADDED_VOCAB_SIZE, BLOCK_SIZE):
+        block_end = min(block_start + BLOCK_SIZE, vocab_size)
+
+        vocab_offset = tl.arange(0, BLOCK_SIZE)
+        mask = vocab_offset < block_end - block_start
+
+        if NO_DRAFT_PROBS:
+            draft_token_id = tl.load(draft_token_ids_ptr + start_idx + pos)
+            prob = tl.load(
+                target_probs_ptr + (start_idx + pos) * vocab_size +
+                block_start + vocab_offset,
+                mask=(mask & (vocab_offset + block_start != draft_token_id)),
+                other=0)
+
+        else:
+            draft_prob = tl.load(draft_probs_ptr +
+                                 (start_idx + pos) * vocab_size + block_start +
+                                 vocab_offset,
+                                 mask=mask,
+                                 other=0)
+            target_prob = tl.load(target_probs_ptr +
+                                  (start_idx + pos) * vocab_size +
+                                  block_start + vocab_offset,
+                                  mask=mask,
+                                  other=0)
+            prob = tl.maximum(target_prob - draft_prob, 0)
+
+            # NOTE(woosuk): We don't need `prob = prob / tl.sum(prob)` here because
+            # `tl.argmax` will select the maximum value.
+
+        q = tl.load(q_ptr + req_idx * vocab_size + block_start + vocab_offset,
+                    mask=mask,
+                    other=float("-inf"))
+
+        # recovered_id = tl.argmax(prob / q, axis=-1)
+        # calc block prob and token ID
+        block_prob = prob / q
+        block_max_prob = tl.max(block_prob, axis=-1)
+        block_best_token_id = tl.argmax(block_prob, axis=-1) + block_start
+
+        # update token ID
+        max_prob = tl.maximum(max_prob, block_max_prob)
+        best_token_id = tl.where(block_max_prob >= max_prob,
+                                 block_best_token_id, best_token_id)
+
+    tl.store(output_token_ids_ptr + start_idx + pos, best_token_id)
+
+
+vllm.v1.sample.rejection_sampler.sample_recovered_tokens_kernel = sample_recovered_tokens_kernel