diff --git a/tensorflow_probability/python/experimental/autobnn/BUILD b/tensorflow_probability/python/experimental/autobnn/BUILD
index 051f58b2ed..6cff94bc4c 100644
--- a/tensorflow_probability/python/experimental/autobnn/BUILD
+++ b/tensorflow_probability/python/experimental/autobnn/BUILD
@@ -56,8 +56,6 @@ py_library(
 py_test(
     name = "bnn_test",
     srcs = ["bnn_test.py"],
-    # TODO(b/320723712): enable this test in OSS.
-    tags = ["no-oss-ci"],
     deps = [
         ":bnn",
         # absl/testing:absltest dep,
@@ -232,12 +230,15 @@ py_library(
 py_test(
     name = "operators_test",
     srcs = ["operators_test.py"],
+    # TODO(b/322864412): enable this test in OSS.
+    tags = ["no-oss-ci"],
     deps = [
         ":kernels",
         ":operators",
         ":util",
         # absl/testing:absltest dep,
         # absl/testing:parameterized dep,
+        # bayeux dep,
         # google/protobuf:use_fast_cpp_protos dep,
         # jax dep,
         # numpy dep,
diff --git a/tensorflow_probability/python/experimental/autobnn/operators_test.py b/tensorflow_probability/python/experimental/autobnn/operators_test.py
index 63f978f003..aab0c34e5b 100644
--- a/tensorflow_probability/python/experimental/autobnn/operators_test.py
+++ b/tensorflow_probability/python/experimental/autobnn/operators_test.py
@@ -15,6 +15,7 @@
 """Tests for operators.py."""
 
 from absl.testing import parameterized
+import bayeux as bx
 import jax
 import jax.numpy as jnp
 import numpy as np
@@ -186,7 +187,6 @@ def test_add_of_adds_has_penultimate(self):
     self.assertEqual((50,), h.shape)
 
   def test_multiply_can_be_trained(self):
-    # TODO(thomaswc): Restore this test when bayeux is available.
     seed = jax.random.PRNGKey(20231018)
     x_train, y_train = util.load_fake_dataset()
 
@@ -202,16 +202,16 @@ def train_density(params):
       return bnn.log_prob(params, x_train, y_train)
 
     transform, inverse_transform, _ = util.make_transforms(bnn)
-    del transform
-    del inverse_transform
-    del init_params
-    del train_density
-    # mix_model = bx.Model(train_density, init_params,
-    #                     transform_fn=transform,
-    #                     inverse_transform_fn=inverse_transform)
-
-    # self.assertTrue(
-    #    mix_model.optimize.optax_adam.debug(seed=seed, verbosity=10))
+    mix_model = bx.Model(
+        train_density,
+        init_params,
+        transform_fn=transform,
+        inverse_transform_fn=inverse_transform,
+    )
+
+    self.assertTrue(
+        mix_model.optimize.optax_adam.debug(seed=seed, verbosity=10)
+    )
 
 
 if __name__ == "__main__":
diff --git a/tensorflow_probability/python/experimental/autobnn/training_util.py b/tensorflow_probability/python/experimental/autobnn/training_util.py
index a1d9d48663..1ff680ce4a 100644
--- a/tensorflow_probability/python/experimental/autobnn/training_util.py
+++ b/tensorflow_probability/python/experimental/autobnn/training_util.py
@@ -17,7 +17,6 @@
 import functools
 from typing import Any, Dict, List, Optional, Sequence, Tuple, Union
 
-import bayeux as bx
 import jax
 import jax.numpy as jnp
 from jaxtyping import PyTree  # pylint: disable=g-importing-member
@@ -38,6 +37,11 @@ def _make_bayeux_model(
     for_vi: bool = False,
 ):
   """Use a MAP estimator to fit a BNN."""
+  # We can't import bayeux at the file level because it would create a
+  # circular dependency:  autobnn imports bayeux imports tfp:jax
+  # which in turn imports (through __init__.py files) autobnn.
+  import bayeux as bx  # pylint:disable=g-bad-import-order,g-import-not-at-top
+
   test_seed, init_seed = jax.random.split(seed)
   test_point = net.init(test_seed, x_train)
   transform, inverse_transform, ildj = util.make_transforms(net)