diff --git a/gigl/src/common/models/pyg/nn/models/feature_embedding.py b/gigl/src/common/models/pyg/nn/models/feature_embedding.py
index 26bc3d645..f98d04021 100644
--- a/gigl/src/common/models/pyg/nn/models/feature_embedding.py
+++ b/gigl/src/common/models/pyg/nn/models/feature_embedding.py
@@ -68,13 +68,12 @@ def __init__(
         self.__padding_idx = padding_idx
         self.__feature_padding_value_map = feature_padding_value_map
 
-        # whether to add 1 to the whole tensor, so all elements in tensor is >= 0 for nn.Embedding input
-        # Since tft.compute_and_apply_vocabulary will be 0 based and use -1 as OOV padding
-        self.__plus_one = False
-        self.__oov_idx: Optional[int] = None
         assert oov_idx is None or oov_idx >= -1, "oov_idx has to be >= -1"
-        if oov_idx and oov_idx == -1:
-            self.__plus_one = True
+        # Per-feature flag: shift indices by +1 when int_domain.min == -1, which is the
+        # single-OOV-bucket case where tft.compute_and_apply_vocabulary uses -1 for OOV.
+        # When num_oov_buckets > 1, TFT assigns OOV indices starting at vocab_size (min == 0),
+        # so no shift is needed — applying +1 would push the max OOV index out of bounds.
+        self.__feature_plus_one: dict[str, bool] = {}
 
         for feature_name, emb_dim in features_to_embed.items():
             feat_dim = get_feature_len_from_fixed_len_feature(
@@ -99,6 +98,7 @@ def __init__(
                 "If int_domain.min_value is -1, oov_idx must also be -1"
             )
             vocab_size = feat_schema.int_domain.max - feat_schema.int_domain.min + 1
+            self.__feature_plus_one[feature_name] = feat_schema.int_domain.min == -1
 
             feature_padding_idx: Optional[int]
             if (
@@ -113,7 +113,7 @@ def __init__(
                 ].index(feature_padding_value)
             else:
                 feature_padding_idx = self.__padding_idx
-            if self.__plus_one and feature_padding_idx is not None:
+            if self.__feature_plus_one[feature_name] and feature_padding_idx is not None:
                 feature_padding_idx = feature_padding_idx + 1
             self.__feature_embedding_layers[feature_name] = nn.Embedding(
                 num_embeddings=vocab_size,
@@ -133,7 +133,7 @@ def forward(self, x: torch.Tensor) -> torch.Tensor:
             x_to_emb = filter_features(
                 feature_schema=self.__feature_schema, feature_names=[feature], x=x
             ).long()  # embedding layer takes LongTensor
-            if self.__plus_one:
+            if self.__feature_plus_one[feature]:
                 x_to_emb = x_to_emb + 1
             emb_layer = self.__feature_embedding_layers[feature]
             emb = emb_layer(x_to_emb)