diff --git a/mteb/encoder_interface.py b/mteb/encoder_interface.py
index 5a66330cdc..abb88f3ac4 100644
--- a/mteb/encoder_interface.py
+++ b/mteb/encoder_interface.py
@@ -179,29 +179,24 @@ def encode(  # current a 1-1 match with Encoder.encode
     ) -> np.ndarray:
         pass
 
-    def get_image_embeddings(  # Seems like sentence transformers use a singular encode for both images and text. Not sure if we want to do the same.
-        # If not it might be ideal to redefine Encoder.encode
+    def get_image_embeddings(
         self,
         images: list[Image.Image] | DataLoader,
         **kwargs,
-        # removed batch_size, it is not required that it will accept kwargs
-    ) -> np.ndarray:  # added standard output (I believe we actually expect tensors in the code, but would like to be consistent)
+    ) -> np.ndarray:
         pass
 
-    def get_text_embeddings(  # any reason for this?
+    def get_text_embeddings(
         self,
         texts: list[str],
         **kwargs,
     ) -> np.ndarray:
         pass
 
-    def get_fused_embeddings(  # hmm what if I have a document with images at specific positions?
+    def get_fused_embeddings(
         self,
         texts: list[str] | None = None,
-        images: list[Image.Image]
-        | DataLoader
-        | None = None,  # the requirement for these two to be the same seems odd (docs without images, images without associated text, docs with multiple images)
-        # fusion_mode: str="sum", # will remove this as it should be required in the interface
+        images: list[Image.Image] | DataLoader | None = None,
         **kwargs: Any,
     ) -> np.ndarray:
         pass