diff --git a/src/lerobot/policies/factory.py b/src/lerobot/policies/factory.py
index b7e8e3a09..79262816b 100644
--- a/src/lerobot/policies/factory.py
+++ b/src/lerobot/policies/factory.py
@@ -36,6 +36,8 @@ from lerobot.processor import (
     transition_to_batch,
     transition_to_policy_action,
 )
+from lerobot.rewards.classifier.configuration_classifier import RewardClassifierConfig
+from lerobot.rewards.sarm.configuration_sarm import SARMConfig
 from lerobot.types import PolicyAction
 from lerobot.utils.constants import (
     ACTION,
@@ -52,8 +54,6 @@ from .pi0.configuration_pi0 import PI0Config
 from .pi05.configuration_pi05 import PI05Config
 from .pretrained import PreTrainedPolicy
 from .sac.configuration_sac import SACConfig
-from .sac.reward_model.configuration_classifier import RewardClassifierConfig
-from .sarm.configuration_sarm import SARMConfig
 from .smolvla.configuration_smolvla import SmolVLAConfig
 from .tdmpc.configuration_tdmpc import TDMPCConfig
 from .utils import validate_visual_features_consistency
@@ -133,7 +133,7 @@ def get_policy_class(name: str) -> type[PreTrainedPolicy]:
 
         return SACPolicy
     elif name == "reward_classifier":
-        from .sac.reward_model.modeling_classifier import Classifier
+        from lerobot.rewards.classifier.modeling_classifier import Classifier
 
         return Classifier
     elif name == "smolvla":
@@ -141,7 +141,7 @@ def get_policy_class(name: str) -> type[PreTrainedPolicy]:
 
         return SmolVLAPolicy
     elif name == "sarm":
-        from .sarm.modeling_sarm import SARMRewardModel
+        from lerobot.rewards.sarm.modeling_sarm import SARMRewardModel
 
         return SARMRewardModel
     elif name == "groot":
@@ -377,7 +377,7 @@ def make_pre_post_processors(
         )
 
     elif isinstance(policy_cfg, RewardClassifierConfig):
-        from .sac.reward_model.processor_classifier import make_classifier_processor
+        from lerobot.rewards.classifier.processor_classifier import make_classifier_processor
 
         processors = make_classifier_processor(
             config=policy_cfg,
@@ -393,7 +393,7 @@ def make_pre_post_processors(
         )
 
     elif isinstance(policy_cfg, SARMConfig):
-        from .sarm.processor_sarm import make_sarm_pre_post_processors
+        from lerobot.rewards.sarm.processor_sarm import make_sarm_pre_post_processors
 
         processors = make_sarm_pre_post_processors(
             config=policy_cfg,