diff --git a/docs/pipes/embeddings/assets/transformer-windowing.svg b/docs/assets/images/transformer-windowing.svg
similarity index 100%
rename from docs/pipes/embeddings/assets/transformer-windowing.svg
rename to docs/assets/images/transformer-windowing.svg
diff --git a/docs/pipeline.md b/docs/pipeline.md
index 8d1d1b00..85d33772 100644
--- a/docs/pipeline.md
+++ b/docs/pipeline.md
@@ -57,7 +57,7 @@ model(pdf_bytes)
 model.pipe([pdf_bytes, ...])
 ```
 
-For more information on how to use the pipeline, refer to the [Inference](../inference) page.
+For more information on how to use the pipeline, refer to the [Inference](/inference) page.
 
 ## Hybrid models
 
diff --git a/edspdf/pipeline.py b/edspdf/pipeline.py
index 8b008d7d..51359064 100644
--- a/edspdf/pipeline.py
+++ b/edspdf/pipeline.py
@@ -854,7 +854,7 @@ def __exit__(ctx_self, type, value, traceback):
         return context()
 
 
-def load(config: Union[Path, str, Config]):
+def load(config: Union[Path, str, Config]) -> Pipeline:
     error = "The load function expects a Config or a path to a config file"
     if isinstance(config, (Path, str)):
         path = Path(config)
diff --git a/edspdf/pipes/embeddings/box_transformer.py b/edspdf/pipes/embeddings/box_transformer.py
index 95f5ac9d..13172544 100644
--- a/edspdf/pipes/embeddings/box_transformer.py
+++ b/edspdf/pipes/embeddings/box_transformer.py
@@ -61,8 +61,8 @@ class BoxTransformer(TrainablePipe[EmbeddingOutput]):
         Initializing with a value close to 0 can help the training converge.
     attention_mode: Sequence[RelativeAttentionMode]
         Mode of relative position infused attention layer.
-        See the [relative attention](relative_attention) documentation for more
-        information.
+        See the [relative attention][edspdf.layers.relative_attention.RelativeAttention]
+        documentation for more information.
     n_layers: int
         Number of layers in the Transformer
     """
diff --git a/edspdf/pipes/embeddings/huggingface_embedding.py b/edspdf/pipes/embeddings/huggingface_embedding.py
index d2d24489..16ccf076 100644
--- a/edspdf/pipes/embeddings/huggingface_embedding.py
+++ b/edspdf/pipes/embeddings/huggingface_embedding.py
@@ -42,7 +42,7 @@ class HuggingfaceEmbedding(TrainablePipe[EmbeddingOutput]):
     occurrence that is the closest to the center of its window.
 
     Here is an overview how this works in a classifier model :
-    ![Transformer windowing](./assets/transformer-windowing.svg)
+    ![Transformer windowing](/assets/images/transformer-windowing.svg)
 
     Examples
     --------
@@ -82,7 +82,8 @@ class HuggingfaceEmbedding(TrainablePipe[EmbeddingOutput]):
     )
     ```
 
-    This model can then be trained following the [training recipe](/recipes/training/).
+    This model can then be trained following the
+    [training recipe](/recipes/training/).
 
     Parameters
     ----------
diff --git a/mkdocs.yml b/mkdocs.yml
index b15ef621..51a1e16f 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -125,6 +125,7 @@ plugins:
             show_root_toc_entry: false
             show_signature: false
             merge_init_into_class: true
+  - autolinks
   - glightbox:
       touchNavigation: true
       loop: false
@@ -155,3 +156,6 @@ markdown_extensions:
   - pymdownx.emoji:
       emoji_index: !!python/name:materialx.emoji.twemoji
       emoji_generator: !!python/name:materialx.emoji.to_svg
+
+validation:
+  absolute_links: ignore
diff --git a/pyproject.toml b/pyproject.toml
index c218fc6d..ea8df287 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -55,6 +55,7 @@ docs = [
     "mkdocstrings~=0.20",
     "mkdocstrings-python~=1.1",
     "mkdocs-autorefs@git+https://github.com/percevalw/mkdocs-autorefs.git@0.4.1.post0",
+    "mkdocs-autolinks-plugin~=0.7.1",
     "mkdocs-gen-files~=0.4.0",
     "mkdocs-literate-nav~=0.6.0",
     "mkdocs-material~=9.1.0",