diff --git a/examples/openarms/evaluate_relative.py b/examples/openarms/evaluate_relative.py
index 7d89ab49c..3c708c0e8 100644
--- a/examples/openarms/evaluate_relative.py
+++ b/examples/openarms/evaluate_relative.py
@@ -26,8 +26,8 @@ from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.pipeline_features import aggregate_pipeline_dataset_features, create_initial_features
 from lerobot.datasets.utils import build_dataset_frame, combine_feature_dicts
 from lerobot.policies.factory import make_policy, make_pre_post_processors
+from lerobot.policies.utils import make_robot_action as tensor_to_robot_action
 from lerobot.processor import make_default_processors
-from lerobot.processor.core import RobotAction
 from lerobot.robots.openarms.config_openarms_follower import OpenArmsFollowerConfig
 from lerobot.robots.openarms.openarms_follower import OpenArmsFollower
 from lerobot.utils.constants import ACTION, OBS_STR
@@ -62,16 +62,6 @@ CAMERA_CONFIG = {
 }
 
 
-def make_robot_action(action_values: dict, features: dict) -> RobotAction:
-    robot_action = {}
-    for key in features:
-        if key.startswith(ACTION + "."):
-            action_key = key.removeprefix(ACTION + ".")
-            if action_key in action_values:
-                robot_action[action_key] = action_values[action_key]
-    return robot_action
-
-
 def load_relative_config(model_path: Path | str) -> tuple[PerTimestepNormalizer | None, bool]:
     """Load normalizer and relative_state setting from checkpoint."""
     model_path = Path(model_path) if isinstance(model_path, str) else model_path
@@ -146,8 +136,8 @@ def inference_loop_relative(
             if isinstance(state_tensor, torch.Tensor):
                 observation_frame[state_key] = convert_state_to_relative(state_tensor)
         
-        # Policy inference (outputs normalized relative actions)
-        action_values = predict_action(
+        # Policy inference (outputs action tensor)
+        action_tensor = predict_action(
             observation=observation_frame,
             policy=policy,
             device=device,
@@ -158,17 +148,20 @@ def inference_loop_relative(
             robot_type=robot.robot_type,
         )
         
-        # Unnormalize actions
+        # Unnormalize relative actions if normalizer exists
         if relative_normalizer is not None:
-            action_keys = [k for k in action_values.keys() if not k.startswith("task")]
-            action_tensor = torch.tensor([[action_values[k] for k in action_keys]])
-            action_tensor = action_tensor.unsqueeze(1)
-            action_unnorm = relative_normalizer.unnormalize(action_tensor)
-            for i, k in enumerate(action_keys):
-                action_values[k] = action_unnorm[0, 0, i].item()
+            # action_tensor shape: [1, action_dim] or [action_dim]
+            if action_tensor.dim() == 1:
+                action_tensor = action_tensor.unsqueeze(0).unsqueeze(0)  # [1, 1, action_dim]
+            elif action_tensor.dim() == 2:
+                action_tensor = action_tensor.unsqueeze(1)  # [batch, 1, action_dim]
+            action_tensor = relative_normalizer.unnormalize(action_tensor)
+            action_tensor = action_tensor.squeeze(1)  # back to [batch, action_dim]
         
-        # Convert to absolute
-        relative_action = make_robot_action(action_values, dataset.features)
+        # Convert tensor to dict
+        relative_action = tensor_to_robot_action(action_tensor, dataset.features)
+        
+        # Convert relative to absolute
         absolute_action = convert_from_relative_actions_dict(relative_action, current_pos)
         
         robot.send_action(absolute_action)