diff --git a/sklearn/linear_model/_least_angle.py b/sklearn/linear_model/_least_angle.py
index 7ed7b27811ec6..35af5dab30267 100644
--- a/sklearn/linear_model/_least_angle.py
+++ b/sklearn/linear_model/_least_angle.py
@@ -69,7 +69,7 @@ def lars_path(
     y : None or array-like of shape (n_samples,)
         Input targets.
 
-    Xy : array-like of shape (n_samples,) or (n_samples, n_targets), \
+    Xy : array-like of shape (n_features,) or (n_features, n_targets), \
             default=None
         `Xy = np.dot(X.T, y)` that can be precomputed. It is useful
         only when the Gram matrix is precomputed.
@@ -215,7 +215,7 @@ def lars_path_gram(
 
     Parameters
     ----------
-    Xy : array-like of shape (n_samples,) or (n_samples, n_targets)
+    Xy : array-like of shape (n_features,) or (n_features, n_targets)
         Xy = np.dot(X.T, y).
 
     Gram : array-like of shape (n_features, n_features)
@@ -362,7 +362,7 @@ def _lars_path_solver(
     y : None or ndarray of shape (n_samples,)
         Input targets.
 
-    Xy : array-like of shape (n_samples,) or (n_samples, n_targets), \
+    Xy : array-like of shape (n_features,) or (n_features, n_targets), \
             default=None
         `Xy = np.dot(X.T, y)` that can be precomputed. It is useful
         only when the Gram matrix is precomputed.
@@ -1110,7 +1110,7 @@ def fit(self, X, y, Xy=None):
         y : array-like of shape (n_samples,) or (n_samples, n_targets)
             Target values.
 
-        Xy : array-like of shape (n_samples,) or (n_samples, n_targets), \
+        Xy : array-like of shape (n_features,) or (n_features, n_targets), \
                 default=None
             Xy = np.dot(X.T, y) that can be precomputed. It is useful
             only when the Gram matrix is precomputed.