diff --git a/drevalpy/datasets/dataset.py b/drevalpy/datasets/dataset.py
index bb46ea8..0613e2e 100644
--- a/drevalpy/datasets/dataset.py
+++ b/drevalpy/datasets/dataset.py
@@ -274,7 +274,7 @@ def split_dataset(
         response = self.response
 
         if mode == "LPO":
-            cv_splits = leave_pair_out_cv(
+            cv_splits = _leave_pair_out_cv(
                 n_cv_splits,
                 response,
                 cell_line_ids,
@@ -287,7 +287,7 @@ def split_dataset(
 
         elif mode in ["LCO", "LDO"]:
             group = "cell_line" if mode == "LCO" else "drug"
-            cv_splits = leave_group_out_cv(
+            cv_splits = _leave_group_out_cv(
                 group=group,
                 n_cv_splits=n_cv_splits,
                 response=response,
@@ -303,7 +303,7 @@ def split_dataset(
 
         if split_validation and split_early_stopping:
             for split in cv_splits:
-                validation_es, early_stopping = split_early_stopping_data(split["validation"], test_mode=mode)
+                validation_es, early_stopping = _split_early_stopping_data(split["validation"], test_mode=mode)
                 split["validation_es"] = validation_es
                 split["early_stopping"] = early_stopping
         self.cv_splits = cv_splits
@@ -444,7 +444,7 @@ def inverse_transform(self, response_transformation: TransformerMixin) -> None:
             self.predictions = response_transformation.inverse_transform(self.predictions.reshape(-1, 1)).squeeze()
 
 
-def split_early_stopping_data(
+def _split_early_stopping_data(
     validation_dataset: DrugResponseDataset, test_mode: str
 ) -> tuple[DrugResponseDataset, DrugResponseDataset]:
     """
@@ -468,7 +468,7 @@ def split_early_stopping_data(
     return validation_dataset, early_stopping_dataset
 
 
-def leave_pair_out_cv(
+def _leave_pair_out_cv(
     n_cv_splits: int,
     response: ArrayLike,
     cell_line_ids: ArrayLike,
@@ -543,7 +543,7 @@ def leave_pair_out_cv(
     return cv_sets
 
 
-def leave_group_out_cv(
+def _leave_group_out_cv(
     group: str,
     n_cv_splits: int,
     response: ArrayLike,