diff --git a/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_cuda_causal_conv_cached_op.py b/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_cuda_causal_conv_cached_op.py
index b2745e51bfd..05ac3c70d22 100644
--- a/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_cuda_causal_conv_cached_op.py
+++ b/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_cuda_causal_conv_cached_op.py
@@ -97,6 +97,7 @@ def test_generate_only_with_slot_mapping_cuda(conv_env):
     )
 
 
+@pytest.mark.skip(reason="https://nvbugspro.nvidia.com/bug/5548861")
 def test_context_flattened_and_state_writeback_cuda(conv_env):
     device = conv_env["device"]
     dtype = conv_env["dtype"]
diff --git a/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_triton_mamba_cached_op.py b/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_triton_mamba_cached_op.py
index 97a23572206..6cce60f1684 100644
--- a/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_triton_mamba_cached_op.py
+++ b/tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_triton_mamba_cached_op.py
@@ -28,6 +28,7 @@ def mamba_env():
     return {"device": device, "dtype": dtype, "atol": atol, "rtol": rtol}
 
 
+@pytest.mark.skip(reason="https://nvbugspro.nvidia.com/bug/5548861")
 def test_triton_generate_only_with_slot_mapping(mamba_env):
     device = mamba_env["device"]
     dtype = mamba_env["dtype"]