diff --git a/neural_tangents/_src/empirical.py b/neural_tangents/_src/empirical.py
index 9009012..1c41abf 100644
--- a/neural_tangents/_src/empirical.py
+++ b/neural_tangents/_src/empirical.py
@@ -125,6 +125,7 @@
 from jax.core import Var
 
 from jax.extend import linear_util as lu
+from jax.extend import source_info_util
 
 from jax.interpreters import ad
 from jax.interpreters.ad import UndefinedPrimal
@@ -1863,7 +1864,7 @@ def read_cotangent(v: Var) -> Union[jnp.ndarray, Zero]:
   map(functools.partial(_write_primal, primal_env), jaxpr.invars, primals_in)
 
   ct_env: dict[Var, jnp.ndarray] = {}
-  ctx = ad.source_info_util.transform_name_stack('transpose')
+  ctx = source_info_util.transform_name_stack('transpose')
   with ctx:
     map(functools.partial(_write_cotangent, 'outvars', ct_env),
         jaxpr.outvars, cotangents_in)
@@ -2174,10 +2175,10 @@ def _eqn_vjp_fn(
     # Identity function
     return cts_in,
 
-  name_stack = (ad.source_info_util.current_name_stack() +
+  name_stack = (source_info_util.current_name_stack() +
                 eqn.source_info.name_stack)
-  with ad.source_info_util.user_context(eqn.source_info.traceback,
-                                        name_stack=name_stack):
+  with source_info_util.user_context(eqn.source_info.traceback,
+                                     name_stack=name_stack):
     if eqn.primitive.call_primitive or eqn.primitive.map_primitive:
       cts_in_avals = [v.aval for v in eqn.outvars]
       params = dict(eqn.params)
@@ -2263,7 +2264,7 @@ def _write_cotangent(
     return
 
   ct_env[v] = ad.add_tangents(ct_env[v], ct) if v in ct_env else ct
-  if ad.config.jax_enable_checks:
+  if jax.config.jax_enable_checks:
     ct_aval = core.get_aval(ct_env[v])
     joined_aval = core.lattice_join(
         v.aval, ct_aval).strip_weak_type().strip_named_shape()