sc-99549-fix-broken-demo-inter-sphinx-links-within-demo (#1550)

## Changes
- Remove `.html` from demo references as we no longer need that in the
URL and its causing 404 with next port, update it to be intersphinx
format

## Ticket
-
[sc-99549-fix-broken-demo-inter-sphinx-links-within-demo](https://app.shortcut.com/xanaduai/story/99549/remove-html-from-demo-references-in-qml-demos)

---------

Co-authored-by: Paul Finlay <[email protected]>
Co-authored-by: Alan Martin <[email protected]>
Co-authored-by: Jonathan Toye <[email protected]>

Author: 4 people

demonstrations_v2/adjoint_diff_benchmarking/metadata.json

@@ -8,7 +8,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2021-11-23T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-09T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Getting Started"
     ],

demonstrations_v2/ahs_aquila/demo.py

@@ -57,7 +57,7 @@
 more information on pulse programming in PennyLane, see the
 `PennyLane docs <https://docs.pennylane.ai/en/stable/code/qml_pulse.html>`__, or check out the demo
 about
-`running a ctrl-VQE algorithm with pulse control <https://pennylane.ai/qml/demos/tutorial_pulse_programming101.html>`__ on the PennyLane `default.qubit` simulator.
+:doc:`running a ctrl-VQE algorithm with pulse control <demos/tutorial_pulse_programming101>` on the PennyLane `default.qubit` simulator.
 
 
 
@@ -66,7 +66,7 @@
 
 The Aquila QPU works with programmable arrays of up to 256 Rubidium-87 atoms (Rb-87), trapped in vacuum by tightly
 focused laser beams. These atoms can be arranged in (almost)
-`arbitrary user-specified geometries <https://pennylane.ai/qml/demos/tutorial_pasqal.html>`_ to determine
+:doc:`arbitrary user-specified geometries <demos/tutorial_pasqal>` to determine
 inter-qubit interactions. On the Aquila device, it is possible to specify 1D and 2D atom arrangements. Atom
 positions may be slightly shifted to accommodate hardware limitations, and must obey lattice constraints
 for spacing. This will be explored in more detail below.

demonstrations_v2/ahs_aquila/metadata.json

@@ -8,7 +8,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2023-05-16T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-09T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Quantum Hardware",
         "Devices and Performance",

demonstrations_v2/braket-parallel-gradients/metadata.json

@@ -14,7 +14,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2020-12-08T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-09T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Devices and Performance"
     ],

demonstrations_v2/circuits_as_fourier_series/metadata.json

@@ -8,7 +8,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2023-09-11T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-17T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Optimization",
         "Quantum Machine Learning"

demonstrations_v2/covalent_cloud_gpu/metadata.json

@@ -11,7 +11,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2024-05-24T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-09T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Quantum Machine Learning"
     ],

demonstrations_v2/ensemble_multi_qpu/demo.py

@@ -46,7 +46,7 @@
 
 ##############################################################################
 # This tutorial requires the ``pennylane-rigetti`` and ``pennylane-qiskit`` packages, which can be
-# installed by following the instructions `here <https://pennylane.ai/install.html>`__. We also
+# installed by following the instructions `here <https://pennylane.ai/install>`__. We also
 # make use of the `PyTorch interface <https://pennylane.readthedocs.io/en/stable/introduction
 # /interfaces.html>`_, which can be installed from `here
 # <https://pytorch.org/get-started/locally/>`__.

demonstrations_v2/ensemble_multi_qpu/metadata.json

@@ -8,7 +8,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2020-02-14T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-09T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Quantum Machine Learning"
     ],

demonstrations_v2/function_fitting_qsp/demo.py

@@ -553,7 +553,7 @@ def step_func(x):
 # explicitly computing the optimal values for :math:`\vec{\phi}`
 # known as "Remez-type exchange algorithms" for analytic function fitting. If
 # you want to explore other approaches to function fitting, checkout this
-# `demo <https://pennylane.ai/qml/demos/quantum_neural_net.html>`__
+# :doc:`demo <demos/quantum_neural_net>`
 # where we use a photonic neural network for function fitting.
 #
 #

demonstrations_v2/function_fitting_qsp/metadata.json

@@ -8,7 +8,7 @@
     "executable_stable": false,
     "executable_latest": false,
     "dateOfPublication": "2022-05-24T00:00:00+00:00",
-    "dateOfLastModification": "2025-09-09T17:41:50+00:00",
+    "dateOfLastModification": "2025-09-22T15:48:14+00:00",
     "categories": [
         "Quantum Machine Learning"
     ],