Humanize concepts docs

docs/conf.py

@@ -107,7 +107,7 @@
     "pygments_dark_style": "monokai",
     "footer_links": {},
     "github_url": "https://github.com/isaac-sim/IsaacLab-Arena",
-    "show_nav_level": 2,
+    "show_nav_level": 1,
     # TODO(alexmillane, 2025-04-24): Try re-enabling this once we have a pypi page.
     # "icon_links": [
     #     {

docs/index.rst

@@ -211,40 +211,26 @@ TABLE OF CONTENTS
    pages/example_workflows/imitation_learning/index
    pages/example_workflows/reinforcement_learning_workflows/index
 
-.. toctree::
-   :maxdepth: 1
-   :caption: Advanced
-
-   pages/advanced/private_omniverse
-   pages/advanced/assets_management
-   pages/advanced/gr00t_server_client
-   pages/quickstart/jupyter_notebooks
-
-.. toctree::
-   :maxdepth: 1
-   :caption: Policy Evaluation
-
-   pages/policy_evaluation/index
-
 .. toctree::
    :maxdepth: 1
    :caption: Concepts
 
    pages/concepts/concept_overview
-   pages/concepts/concept_environment_design
-   pages/concepts/concept_embodiment_design
-   pages/concepts/concept_tasks_design
-   pages/concepts/concept_sequential_tasks_design
-   pages/concepts/concept_scene_design
-   pages/concepts/concept_metrics_design
-   pages/concepts/concept_teleop_devices_design
    pages/concepts/concept_environment_compilation
-   pages/concepts/concept_assets_design
-   pages/concepts/concept_affordances_design
+   pages/concepts/scene/index
+   pages/concepts/task/index
+   pages/concepts/embodiment/index
    pages/concepts/concept_object_placement
-   pages/concepts/concept_policy_design
-   pages/concepts/concept_remote_policies_design
+   pages/concepts/policy/index
+
+.. toctree::
+   :maxdepth: 1
+   :caption: Advanced
 
+   pages/advanced/private_omniverse
+   pages/advanced/assets_management
+   pages/advanced/gr00t_server_client
+   pages/quickstart/jupyter_notebooks
 
 .. toctree::
    :maxdepth: 1

docs/pages/advanced/gr00t_server_client.rst

@@ -3,7 +3,7 @@ Server-Client Mode With GR00T
 
 Arena supports running the simulation and the policy model in separate containers connected via a
 lightweight RPC protocol. This is useful when the policy environment has different dependencies or
-needs to run on a different machine. You can see more details in the :doc:`../concepts/concept_remote_policies_design` page.
+needs to run on a different machine. You can see more details in the :doc:`../concepts/policy/concept_remote_policies_design` page.
 
 It requires two containers to run:
 
@@ -59,4 +59,4 @@ The simulation environment and the policy model communicate via the remote polic
 If you want to host other policy models as remote servers, you can follow the same pattern: create
 a dedicated server Dockerfile and launcher script (similar to ``docker/Dockerfile.gr00t_server``
 and ``docker/run_gr00t_server.sh``), and point it to a custom ``ServerSidePolicy``
-implementation as described in :doc:`../concepts/concept_remote_policies_design`.
+implementation as described in :doc:`../concepts/policy/concept_remote_policies_design`.