diff --git a/pytensor/link/pytorch/dispatch/basic.py b/pytensor/link/pytorch/dispatch/basic.py
index 449b9ee779..6faf38ae0d 100644
--- a/pytensor/link/pytorch/dispatch/basic.py
+++ b/pytensor/link/pytorch/dispatch/basic.py
@@ -4,6 +4,7 @@
 import torch
 import torch.compiler
 
+from pytensor.compile import PYTORCH
 from pytensor.compile.builders import OpFromGraph
 from pytensor.compile.ops import DeepCopyOp
 from pytensor.graph.fg import FunctionGraph
@@ -137,25 +138,13 @@ def makevector(*x):
 
 
 @pytorch_funcify.register(OpFromGraph)
-def pytorch_funcify_OpFromGraph(op, node=None, **kwargs):
-    _ = kwargs.pop("storage_map", None)
+def pytorch_funcify_OpFromGraph(op, node, **kwargs):
+    kwargs.pop("storage_map", None)
 
-    fgraph_fn = torch.compile(pytorch_funcify(op.fgraph, **kwargs))
+    # Apply inner rewrites
+    PYTORCH.optimizer(op.fgraph)
 
-    if len(op.fgraph.outputs) > 1:
-
-        def inner(*args):
-            return fgraph_fn(*args)
-    else:
-
-        def inner(*args):
-            return fgraph_fn(*args)[0]
-
-    # Don't compile the inner function
-    # This is due torch failing to create
-    # guards when parent scoped closure variables
-    # are used in conditional statements.
-    # Instead of rewriting many portions of code
-    # this will allow for only this small section to
-    # not be compiled by the outer graph
-    return torch.compiler.disable(inner)
+    fgraph_fn = pytorch_funcify(op.fgraph, **kwargs, squeeze_output=True)
+    # Disable one step inlining to prevent torch from trying to import local functions
+    # defined in `pytorch_funcify`
+    return torch.compiler.disable(fgraph_fn, recursive=False)
diff --git a/tests/link/pytorch/test_basic.py b/tests/link/pytorch/test_basic.py
index 4b0790a129..1945790fda 100644
--- a/tests/link/pytorch/test_basic.py
+++ b/tests/link/pytorch/test_basic.py
@@ -68,9 +68,9 @@ def compare_pytorch_and_py(
 
     if len(fgraph.outputs) > 1:
         for j, p in zip(pytorch_res, py_res):
-            assert_fn(j.cpu(), p)
+            assert_fn(j.detach().cpu().numpy(), p)
     else:
-        assert_fn([pytorch_res[0].cpu()], py_res)
+        assert_fn(pytorch_res[0].detach().cpu().numpy(), py_res[0])
 
     return pytensor_torch_fn, pytorch_res