diff --git a/doc/api.rst b/doc/api.rst
index fabd7b6a..e2216653 100644
--- a/doc/api.rst
+++ b/doc/api.rst
@@ -287,6 +287,7 @@ Top-level classes and functions
 
 .. autoclass:: genno.Quantity
    :members:
+   :inherited-members: pipe, shape, size
 
 The :class:`.Quantity` constructor converts its arguments to an internal, :class:`xarray.DataArray`-like data format:
 
@@ -424,3 +425,6 @@ Utilities for testing
 .. automodule:: genno.testing
    :members:
    :exclude-members: parametrize_quantity_class
+
+.. automodule:: genno.testing.jupyter
+   :members:
diff --git a/doc/cache.rst b/doc/cache.rst
index 8b8bbcdd..4d4d0de8 100644
--- a/doc/cache.rst
+++ b/doc/cache.rst
@@ -31,7 +31,7 @@ A cache key is computed from:
 
 1. the name of `func`.
 2. the arguments to `func`, and
-3. the compiled bytecode of `func` (see :func:`hash_code`).
+3. the compiled bytecode of `func` (see :func:`.hash_code`).
 
 If a file exists in ``cache_path`` with a matching key, it is loaded and returned instead of calling `func`.
 
@@ -61,12 +61,12 @@ Consider a function that loads a very large file, or performs some slow processi
         # … further processing …
         return Quantity(result)
 
-We want to cache the result of :func:`slow_data_load`, but have the cache refreshed when the file contents change.
+We want to cache the result of :py:`slow_data_load`, but have the cache refreshed when the file contents change.
 We do this using the `_extra_cache_key` argument to the function.
 This argument is not used in the function, but *does* affect the value of the cache key.
 
 When calling the function, pass some value that indicates whether the contents of `path` have changed.
-One possibility is the modification time, via :meth:`.Path.stat`:
+One possibility is the modification time, via :meth:`pathlib.Path.stat`:
 
 .. code-block:: python
 
@@ -74,7 +74,7 @@ One possibility is the modification time, via :meth:`.Path.stat`:
         return slow_data_load(path, path.stat().st_mtime)
 
 Another possibility is to hash the entire file.
-:func:`hash_contents` is provided for this purpose:
+:func:`.hash_contents` is provided for this purpose:
 
 .. code-block:: python
 
diff --git a/doc/compat-pyam.rst b/doc/compat-pyam.rst
index 85805567..7de5b1a5 100644
--- a/doc/compat-pyam.rst
+++ b/doc/compat-pyam.rst
@@ -45,10 +45,8 @@ Pyam (:mod:`.compat.pyam`)
 
       as_pyam
       add_as_pyam
-      concat
-      write_report
 
-   .. .. autofunction:: as_pyam
+   This module also registers implementations of :func:`.concat` and :func:`.write_report` that handle :class:`pyam.IamDataFrame` objects.
 
    .. autofunction:: add_as_pyam
 
diff --git a/doc/index.rst b/doc/index.rst
index 05f7f6fa..4ae40554 100644
--- a/doc/index.rst
+++ b/doc/index.rst
@@ -4,7 +4,7 @@
 **genno** is a Python package for describing and executing complex calculations on labelled, multi-dimensional data.
 It aims to make these calculations efficient, transparent, and easily validated as part of scientific research.
 
-genno is built on high-quality Python data packages including ``dask``, ``xarray``, and ``pint``; and provides (current or planned) compatibility with packages including ``pandas``, ``matplotlib``, ``plotnine``, ``ixmp``, ``sdmx1``, and ``pyam``.
+genno is built on high-quality Python data packages including :py:`dask`, :mod:`xarray`, and ```pint``; and provides (current or planned) compatibility with packages including :mod:`pandas`, :mod:`matplotlib`, :mod:`plotnine <.compat.plotnine>`, :mod:`ixmp`, :mod:`sdmx1 <.compat.sdmx>`, and :mod:`pyam <.compat.pyam>`.
 
 .. toctree::
    :maxdepth: 2
@@ -22,9 +22,9 @@ Compatibility
 
 :mod:`.genno` provides built-in support for interaction with:
 
-- :doc:`Plotnine <compat-plotnine>` (:mod:`.plotnine`), via :mod:`.compat.plotnine`.
-- :doc:`Pyam <compat-pyam>` (:mod:`.pyam`), via :mod:`.compat.pyam`.
-- :doc:`SDMX <compat-sdmx>` (:mod:`.sdmx`), via :mod:`.compat.sdmx`.
+- :doc:`Plotnine <compat-plotnine>` (:mod:`plotnine`), via :mod:`.compat.plotnine`.
+- :doc:`Pyam <compat-pyam>` (:mod:`pyam`), via :mod:`.compat.pyam`.
+- :doc:`SDMX <compat-sdmx>` (:mod:`sdmx`), via :mod:`.compat.sdmx`.
 
 .. toctree::
    :maxdepth: 1
@@ -37,8 +37,8 @@ Compatibility
 
 Packages that extend :mod:`genno` include:
 
-- :mod:`ixmp.reporting`
-- :mod:`message_ix.reporting`
+- :mod:`ixmp.report`
+- :mod:`message_ix.report`
 
 .. toctree::
    :maxdepth: 2
diff --git a/doc/whatsnew.rst b/doc/whatsnew.rst
index 0f109f76..3ebf0048 100644
--- a/doc/whatsnew.rst
+++ b/doc/whatsnew.rst
@@ -123,7 +123,7 @@ v1.17.0 (2023-05-15)
 ====================
 
 - Bug fix: genno v1.16.1 (:pull:`85`) introduced :class:`ValueError` for some usages of :func:`.computations.sum <.operator.sum>` (:pull:`88`).
-- Provide typed signatures for :meth:`Quantity.bfill`, :meth:`~Quantity.cumprod`, :meth:`~Quantity.ffill` for the benefit of downstream applications (:pull:`88`).
+- Provide typed signatures for :meth:`.Quantity.bfill`, :meth:`~.Quantity.cumprod`, :meth:`~.Quantity.ffill` for the benefit of downstream applications (:pull:`88`).
 - Ensure and test that :attr:`.Quantity.name` and :attr:`~.Quantity.units` pass through all :mod:`~genno.computations <genno.operator>`, in particular :func:`~.operator.aggregate`, :func:`~.operator.convert_units`, and :func:`~.operator.sum` (:pull:`88`).
 - Simplify arithmetic operations (:func:`~.operator.div`, :func:`~.operator.mul`, :func:`~.operator.pow`) so they are agnostic as to the :class:`.Quantity` class in use (:pull:`88`).
 - Ensure :attr:`.AttrSeries.index` is always :class:`pandas.MultiIndex` (:pull:`88`).
@@ -144,13 +144,13 @@ v1.16.0 (2023-04-29)
 - genno supports and is tested on Python 3.11 (:pull:`83`).
 - Update dependencies (:pull:`83`):
 
-  - General: :mod:`importlib_resources` (the independent backport of :mod:`importlib.resources`) is added for Python 3.9 and earlier.
+  - General: :py:`importlib_resources` (the independent backport of :mod:`importlib.resources`) is added for Python 3.9 and earlier.
   - ``genno[sparse]``: new set of optional dependencies, including :mod:`sparse`.
     Install this set in order to use :class:`.SparseDataArray` for :class:`.Quantity`.
 
-    Note that sparse depends on :mod:`numba`, and thus :mod:`llvmlite`, and both of these package can lag new Python versions by several months.
+    Note that sparse depends on :py:`numba`, and thus :py:`llvmlite`, and both of these package can lag new Python versions by several months.
     For example, as of this release, they do not yet support Python 3.11, and thus :mod:`sparse` and :class:`.SparseDataArray` can only be used with Python 3.10 and earlier.
-  - ``genno[tests]``: :mod:`ixmp` is removed; :mod:`jupyter` and :mod:`nbclient` are added.
+  - ``genno[tests]``: :mod:`ixmp` is removed; :py:`jupyter` and :py:`nbclient` are added.
     Testing utilities in :mod:`genno.testing.jupyter` are duplicated from :mod:`ixmp.testing.jupyter`.
 
 - Adjust :meth:`.AttrSeries.interp` for compatibility with pandas 2.0.0 (released 2023-04-03) (:pull:`81`).
@@ -268,7 +268,7 @@ v1.9.1 (2022-01-27)
 Note that installing ``genno[pyam]`` (including via ``genno[compat]``) currently forces the installation of an old version of :mod:`pint`; version 0.17 or earlier.
 Users wishing to use :mod:`genno.compat.pyam` should first install ``genno[pyam]``, then ``pip install --upgrade pint`` to restore a recent version of pint (0.18 or newer) that is usable with genno.
 
-- :func:`computations.concat` works with :class:`.AttrSeries` with misaligned dimensions (:pull:`53`).
+- :func:`.computations.concat <.operator.concat>` works with :class:`.AttrSeries` with misaligned dimensions (:pull:`53`).
 - Improve typing of :class:`.Quantity` and :class:`.Computer` to help with using `mypy <https://mypy.readthedocs.io>`_ on code that uses :mod:`genno` (:pull:`53`).
 
 v1.9.0 (2021-11-23)
@@ -307,7 +307,7 @@ v1.7.0 (2021-07-22)
 v1.6.0 (2021-07-07)
 ===================
 
-- Add :meth:`Key.permute_dims` (:pull:`47`).
+- Add :py:`Key.permute_dims()` (:pull:`47`).
 - Improve performance of :meth:`.Computer.check_keys` (:pull:`47`).
 
 v1.5.2 (2021-07-06)
@@ -323,9 +323,9 @@ v1.5.1 (2021-07-01)
 v1.5.0 (2021-06-27)
 ===================
 
-- Adjust :meth:`.test_assign_coords` for xarray 0.18.2 (:pull:`43`).
+- Adjust :func:`.test_assign_coords` for xarray 0.18.2 (:pull:`43`).
 - Make :attr:`.Key.dims` order-insensitive so that ``Key("foo", "ab") == Key("foo", "ba")`` (:pull:`42`); make corresponding changes to :class:`.Computer` (:pull:`44`).
-- Fix “:class:`AttributeError`: 'COO' object has no attribute 'item'” on :meth:`SparseDataArray.item` (:pull:`41`).
+- Fix “:class:`AttributeError`: 'COO' object has no attribute 'item'” on :meth:`.SparseDataArray.item` (:pull:`41`).
 
 v1.4.0 (2021-04-26)
 ===================
@@ -339,18 +339,18 @@ v1.3.0 (2021-03-22)
 
 - Bump minimum version of :mod:`sparse` from 0.10 to 0.12 and adjust to changes in this version (:pull:`39`)
 
-  - Remove :meth:`.SparseDataArray.equals`, obviated by improvements in :mod:`sparse`.
+  - Remove :py:`SparseDataArray.equals()`, obviated by improvements in :mod:`sparse`.
 
 - Improve :class:`.AttrSeries` (:pull:`39`)
 
   - Implement :meth:`~.AttrSeries.drop_vars` and :meth:`~.AttrSeries.expand_dims`.
   - :meth:`~.AttrSeries.assign_coords` can relabel an entire dimension.
-  - :meth:`~.AttrSeries.sel` can accept :class:`.DataArray` indexers and rename/combine dimensions.
+  - :meth:`~.AttrSeries.sel` can accept :class:`~xarray.DataArray` indexers and rename/combine dimensions.
 
 v1.2.1 (2021-03-08)
 ===================
 
-- Bug fix: Provide abstract :class:`.Quantity.to_series` method for type checking in packages that depend on :mod:`genno`.
+- Bug fix: Provide abstract :meth:`.Quantity.to_series` method for type checking in packages that depend on :mod:`genno`.
 
 v1.2.0 (2021-03-08)
 ===================
@@ -377,15 +377,15 @@ v1.1.0 (2021-02-16)
 v1.0.0 (2021-02-13)
 ===================
 
-- Adjust for usage by :mod:`ixmp.reporting` and :mod:`message_ix.reporting` (:pull:`28`):
+- Adjust for usage by :mod:`ixmp.reporting <ixmp.report>` and :mod:`message_ix.reporting <message_ix.report>` (:pull:`28`):
 
   - Reduce minimum Python version to 3.6.
     This is lower than the minimum version for xarray (3.7), but matches ixmp, etc.
-  - Remove :mod:`compat.ixmp`; this code has been moved to :mod:`ixmp.reporting`, replacing what was there.
-    Likewise, remove :mod:`compat.message_ix`.
+  - Remove submodule :py:`compat.ixmp`; this code has been moved to :mod:`ixmp.reporting <ixmp.report>`, replacing what was there.
+    Likewise, remove submodule :py:`compat.message_ix`.
   - Simplify the form & parsing of ``iamc:`` section entries in configuration files:
 
-    - Remove unused feature to add :func:`group_sum` to the chain of tasks.
+    - Remove unused feature to add :py:`group_sum()` to the chain of tasks.
     - Keys now conform more closely to the arguments of :meth:`.Computer.convert_pyam`.
 
   - Move argument-checking from :func:`.as_pyam` to :meth:`.convert_pyam()`.
@@ -416,11 +416,11 @@ v0.2.0 (2021-01-18)
 -------------------
 
 - Increase test coverage to 100% (:pull:`12`).
-- Port code from :mod:`message_ix.reporting` (:pull:`11`).
+- Port code from :mod:`message_ix.reporting <message_ix.report>` (:pull:`11`).
 - Add :mod:`.compat.pyam`.
 - Add a `name` parameter to :func:`.load_file`.
 
 v0.1.0 (2021-01-10)
 -------------------
 
-- Initial code port from :mod:`ixmp.reporting`.
+- Initial code port from :mod:`ixmp.reporting <ixmp.report>`.
diff --git a/genno/caching.py b/genno/caching.py
index 1fa0a2e4..0481f1f0 100644
--- a/genno/caching.py
+++ b/genno/caching.py
@@ -53,7 +53,7 @@ def ignore(cls, *types):
 
     @classmethod
     def register(cls, func):
-        """Register a `func` to serialize a type not handled by :class:`.JSONEncoder`.
+        """Register `func` to serialize a type not handled by :class:`json.JSONEncoder`.
 
         `func` should return a type that *is* handled by JSONEncoder; see the docs.
 
@@ -69,12 +69,12 @@ def register(cls, func):
         return _encode.register(func)
 
     def default(self, o):
-        """For `o`, return an object serializable by the base :class:`.JSONEncoder`.
+        """For `o`, return an object serializable by the base :class:`json.JSONEncoder`.
 
         - :class:`pathlib.Path`: the string representation of `o`.
-        - :class:`code` objects (from Python's built-in :mod:`inspect` module), e.g.
-          a function or lambda: :class:`blake2b` hash of the object's bytecode and its
-          serialized constants.
+        - :ref:`python:code-objects` (from Python's built-in :mod:`inspect` module), for
+          instance a function or lambda: :func:`~hashlib.blake2b` hash of the object's
+          bytecode and its serialized constants.
 
           .. warning:: This is not 100% guaranteed to change if the operation of `o` (or
              other code called in turn by `o`) changes. If relying on this behaviour,
@@ -102,7 +102,7 @@ def default(self, o):
 
 
 def hash_args(*args, **kwargs):
-    """Return a 20-character :class:`blake2b` hex digest of `args` and `kwargs`.
+    """Return a 20-character :func:`hashlib.blake2b` hex digest of `args` and `kwargs`.
 
     Used by :func:`.decorate`.
 
@@ -121,7 +121,7 @@ def hash_args(*args, **kwargs):
 
 
 def hash_code(func: Callable) -> str:
-    """Return the :class:`.blake2b` hex digest of the compiled bytecode of `func`.
+    """Return the :func:`hashlib.blake2b` hex digest of the compiled bytecode of `func`.
 
     See also
     --------
@@ -133,11 +133,11 @@ def hash_code(func: Callable) -> str:
 
 
 def hash_contents(path: Union[Path, str], chunk_size=65536) -> str:
-    """Return the :class:`.blake2b` hex digest of the contents of the file at `path`.
+    """Return the :func:`hashlib.blake2b` hex digest the file contents of `path`.
 
     Parameters
     ----------
-    chunk_size : int, *optional*
+    chunk_size : int, optional
         Read the file in chunks of this size; default 64 kB.
     """
     with Path(path).open("rb") as f:
@@ -154,13 +154,13 @@ def decorate(
 
     Parameters
     ----------
-    computer : Computer, *optional*
+    computer : .Computer, optional
         If supplied, the ``config`` dictionary stored in the Computer is used to look
         up values for `cache_path` and `cache_skip`, at the moment when `func` is
         called.
-    cache_path : os.Pathlike, *optional*
+    cache_path : os.PathLike, optional
         Directory in which to store cache files.
-    cache_skip : bool, *optional*
+    cache_skip : bool, optional
         If :obj:`True`, ignore existing cache entries and overwrite them with new
         values from `func`.
 
diff --git a/genno/compat/graphviz.py b/genno/compat/graphviz.py
index 4e161eee..6906eaf0 100644
--- a/genno/compat/graphviz.py
+++ b/genno/compat/graphviz.py
@@ -51,11 +51,11 @@ def visualize(
     ----------
     dsk :
         The graph to display.
-    filename : Path or str, *optional*
+    filename : Path or str, optional
         The name of the file to write to disk. If the file name does not have a suffix,
         ".png" is used by default. If `filename` is :data:`None`, no file is written,
         and dask communicates with :program:`dot` using only pipes.
-    format : {'png', 'pdf', 'dot', 'svg', 'jpeg', 'jpg'}, *optional*
+    format : {'png', 'pdf', 'dot', 'svg', 'jpeg', 'jpg'}, optional
         Format in which to write output file, if not given by the suffix of `filename`.
         Default "png".
     data_attributes :
@@ -73,7 +73,7 @@ def visualize(
     edge_attr :
         Mapping of (attribute, value) pairs set for all edges. Passed directly to
         :class:`.graphviz.Digraph`.
-    collapse_outputs : bool, *optional*
+    collapse_outputs : bool, optional
         Omit nodes for keys that are the output of intermediate calculations.
     kwargs :
         All other keyword arguments are added to `graph_attr`.
diff --git a/genno/compat/plotnine/plot.py b/genno/compat/plotnine/plot.py
index dba5792c..b7377197 100644
--- a/genno/compat/plotnine/plot.py
+++ b/genno/compat/plotnine/plot.py
@@ -1,21 +1,19 @@
 import logging
 from abc import ABC, abstractmethod
-from typing import TYPE_CHECKING, Hashable, Sequence
+from typing import Hashable, Sequence
 from warnings import warn
 
 import plotnine as p9
 
+from genno.core.computer import Computer
+from genno.core.key import KeyLike
 from genno.core.quantity import Quantity
 
-if TYPE_CHECKING:
-    from genno.core.computer import Computer
-    from genno.core.key import KeyLike
-
 log = logging.getLogger(__name__)
 
 
 class Plot(ABC):
-    """Class for plotting using :mod:`plotnine`."""
+    """Class for plotting using :doc:`plotnine <plotnine:index>`."""
 
     #: Filename base for saving the plot.
     basename = ""
@@ -84,7 +82,7 @@ def make_task(cls, *inputs):
 
         Parameters
         ----------
-        inputs : sequence of :class:`.Key`, :class:`str`, or other hashable, *optional*
+        *inputs : `.Key` or str or hashable, optional
             If provided, overrides the :attr:`inputs` property of the class.
 
         Returns
@@ -106,8 +104,8 @@ def make_task(cls, *inputs):
 
     @classmethod
     def add_tasks(
-        cls, c: "Computer", key: "KeyLike", *inputs, strict: bool = False
-    ) -> "KeyLike":
+        cls, c: Computer, key: KeyLike, *inputs, strict: bool = False
+    ) -> KeyLike:
         """Add a task to `c` to generate and save the Plot.
 
         Analogous to :meth:`.Operator.add_tasks`.
@@ -125,8 +123,8 @@ def generate(self, *args, **kwargs):
 
         Parameters
         ----------
-        args : sequence of :class:`pandas.DataFrame`
-            Because :mod:`plotnine` operates on pandas data structures, :meth:`save`
-            automatically converts :obj:`Quantity` before being provided to
-            :meth:`generate`.
+        args : sequence of pandas.DataFrame
+            Because :doc:`plotnine <plotnine:index>` operates on pandas data structures,
+            :meth:`save` automatically converts :obj:`.Quantity` before they are passed
+            to :meth:`generate`.
         """
diff --git a/genno/compat/pyam/operator.py b/genno/compat/pyam/operator.py
index 597b94a1..04ecf93a 100644
--- a/genno/compat/pyam/operator.py
+++ b/genno/compat/pyam/operator.py
@@ -1,7 +1,15 @@
 import logging
 from functools import partial
 from pathlib import Path
-from typing import TYPE_CHECKING, Callable, Collection, Iterable, Optional, Union
+from typing import (
+    TYPE_CHECKING,
+    Callable,
+    Collection,
+    Iterable,
+    Mapping,
+    Optional,
+    Union,
+)
 from warnings import warn
 
 import pyam
@@ -27,7 +35,7 @@ def as_pyam(
     scenario,
     quantity: "Quantity",
     *,
-    rename=dict(),
+    rename=Mapping[str, str],
     collapse: Optional[Callable] = None,
     replace=dict(),
     drop: Union[Collection[str], str] = "auto",
@@ -54,22 +62,24 @@ def as_pyam(
     Parameters
     ----------
     scenario :
-        Any object with :attr:`model` and :attr:`scenario` attributes of type
-        :class:`str`, for instance an :class:`ixmp.Scenario`.
-    rename : dict (str -> str), *optional*
-        Mapping from dimension names in `quantity` to column names; either IAMC
-        dimension names, or others that are consumed by `collapse`.
-    collapse : callable, *optional*
+        Any object with :py:`model` and :py:`scenario` attributes of type :class:`str`,
+        for instance an :class:`ixmp.Scenario` or
+        :class:`~message_ix_models.util.scenarioinfo.ScenarioInfo`.
+    rename : dict, optional
+        Mapping from dimension names in `quantity` (:class:`str`) to column names
+        (:class:`str`); either IAMC dimension names, or others that are consumed by
+        `collapse`.
+    collapse : callable, optional
         Function that takes a :class:`pandas.DataFrame` and returns the same type.
         This function **may** collapse 2 or more dimensions, for example to construct
         labels for the IAMC ``variable`` dimension, or any other.
-    replace : *optional*
+    replace : optional
         Values to be replaced and their replaced. Passed directly to
         :meth:`pandas.DataFrame.replace`.
-    drop : str or collection of str, *optional*
+    drop : str or collection of str, optional
         Columns to drop. Passed to :func:`.util.drop`, so if not given, all non-IAMC
         columns are dropped.
-    unit : str, *optional*
+    unit : str, optional
         Label for the IAMC ``unit`` dimension. Passed to
         :func:`~.pyam.util.clean_units`.
 
@@ -130,9 +140,9 @@ def add_as_pyam(
 
     Parameters
     ----------
-    quantities : str or Key or list of (str, Key)
+    quantities : str or .Key or list of str or .Key
         Keys for quantities to transform.
-    tag : str, *optional*
+    tag : str, optional
         Tag to append to new Keys.
 
     Other parameters
@@ -142,7 +152,7 @@ def add_as_pyam(
 
     Returns
     -------
-    list of Key
+    list of .Key
         Each task converts a :class:`.Quantity` into a :class:`pyam.IamDataFrame`.
     """
     # Handle single vs. iterable of inputs
diff --git a/genno/compat/sdmx.py b/genno/compat/sdmx.py
index d904cd59..21f6696d 100644
--- a/genno/compat/sdmx.py
+++ b/genno/compat/sdmx.py
@@ -1,11 +1,12 @@
 from typing import TYPE_CHECKING, Iterable, List, Mapping, Optional, Union
 
 if TYPE_CHECKING:
-    from sdmx.model.common import Code, Codelist
+    import sdmx.model.common
 
 
 def codelist_to_groups(
-    codes: Union["Codelist", Iterable["Code"]], dim: Optional[str] = None
+    codes: Union["sdmx.model.common.Codelist", Iterable["sdmx.model.common.Code"]],
+    dim: Optional[str] = None,
 ) -> Mapping[str, Mapping[str, List[str]]]:
     """Convert `codes` into a mapping from parent items to their children.
 
@@ -16,14 +17,14 @@ def codelist_to_groups(
     codes
         Either a :class:`sdmx.Codelist <sdmx.model.common.Codelist>` object or any
         iterable of :class:`sdmx.Code <sdmx.model.common.Code>`.
-    dim : str, *optional*
+    dim : str, optional
         Dimension to aggregate. If `codes` is a code list and `dim` is not given, the
         ID of the code list is used; otherwise `dim` must be supplied.
     """
     from sdmx.model.common import Codelist
 
     if isinstance(codes, Codelist):
-        items: Iterable["Code"] = codes.items.values()
+        items: Iterable["sdmx.model.common.Code"] = codes.items.values()
         dim = dim or codes.id
     else:
         items = codes
diff --git a/genno/compat/xarray.py b/genno/compat/xarray.py
index 7cf8331e..564c1a52 100644
--- a/genno/compat/xarray.py
+++ b/genno/compat/xarray.py
@@ -85,6 +85,7 @@ def attrs(self) -> Dict[Any, Any]:
 
     @property
     def data(self) -> Any:
+        """Like :attr:`xarray.DataArray.data`."""
         return NotImplemented
 
     @property
@@ -99,10 +100,12 @@ def dims(self) -> Tuple[Hashable, ...]:
 
     @property
     def shape(self) -> Tuple[int, ...]:
+        """Like :attr:`xarray.DataArray.shape`."""
         return NotImplemented
 
     @property
     def size(self) -> int:
+        """Like :attr:`xarray.DataArray.size`."""
         return NotImplemented
 
     def assign_coords(
@@ -122,6 +125,7 @@ def astype(
         copy=None,
         keep_attrs=True,
     ):
+        """Like :meth:`xarray.DataArray.astype`."""
         ...
 
     def bfill(
@@ -129,6 +133,7 @@ def bfill(
         dim: Hashable,
         limit: Optional[int] = None,
     ):
+        """Like :meth:`xarray.DataArray.bfill`."""
         ...
 
     def copy(
@@ -146,6 +151,7 @@ def cumprod(
         keep_attrs: Optional[bool] = None,
         **kwargs: Any,
     ):
+        """Like :meth:`xarray.DataArray.cumprod`."""
         ...
 
     def drop_vars(
@@ -169,6 +175,7 @@ def ffill(
         dim: Hashable,
         limit: Optional[int] = None,
     ):
+        """Like :meth:`xarray.DataArray.ffill`."""
         return NotImplemented
 
     def groupby(
@@ -198,6 +205,7 @@ def pipe(
         *args: Any,
         **kwargs: Any,
     ) -> T:
+        """Like :meth:`xarray.DataArray.pipe`."""
         return NotImplemented
 
     def rename(
@@ -226,6 +234,7 @@ def shift(
         fill_value: Any = None,
         **shifts_kwargs: int,
     ):
+        """Like :attr:`xarray.DataArray.shift`."""
         ...
 
     def sum(
diff --git a/genno/config.py b/genno/config.py
index 3e323163..7e158d9c 100644
--- a/genno/config.py
+++ b/genno/config.py
@@ -43,7 +43,7 @@ def configure(path: Optional[Union[Path, str]] = None, **config):
 
     Parameters
     ----------
-    path : pathlib.Path, *optional*
+    path : pathlib.Path, optional
         Path to a configuration file in JSON or YAML format.
     **config :
         Configuration keys/sections and values.
@@ -62,11 +62,11 @@ def handles(section_name: str, iterate: bool = True, discard: bool = True):
     section_name: str
         The name of the configuration section to handle. Using a name already present
         in :data:`HANDLERS` overrides that handler.
-    iterate : bool, *optional*
+    iterate : bool, optional
         If :obj:`True`, the handler is called once for each item (either list item, or
         (key, value) tuple) in the section. If :obj:`False`, the entire section
         contents, of whatever type, are passed to tha handler.
-    discard : bool, *optional*
+    discard : bool, optional
         If :obj:`True`, configuration section data is discarded after the handler is
         called. If :obj:`False`, the data is retained and stored on the Computer.
     """
@@ -264,6 +264,8 @@ def files(c: Computer, info):
 def general(c: Computer, info):
     """Handle one entry from the ``general:`` config section."""
     # Inputs
+    # TODO allow to specify a more narrow key and *not* have infer_keys applied; perhaps
+    #      using "*"
     inputs = c.infer_keys(info.get("inputs", []))
 
     if info["comp"] in ("mul", "product"):
diff --git a/genno/core/attrseries.py b/genno/core/attrseries.py
index 29d13705..27c86c5b 100644
--- a/genno/core/attrseries.py
+++ b/genno/core/attrseries.py
@@ -99,10 +99,10 @@ class AttrSeries(pd.Series, Quantity):
 
     Parameters
     ----------
-    units : str or pint.Unit, *optional*
+    units : str or pint.Unit, optional
         Set the units attribute. The value is converted to :class:`pint.Unit` and added
         to `attrs`.
-    attrs : :class:`~collections.abc.Mapping`, *optional*
+    attrs : :class:`~collections.abc.Mapping`, optional
         Set the :attr:`~pandas.Series.attrs` of the AttrSeries. This attribute was added
         in `pandas 1.0 <https://pandas.pydata.org/docs/whatsnew/v1.0.0.html>`_, but is
         not currently supported by the Series constructor.
diff --git a/genno/core/computer.py b/genno/core/computer.py
index 608a6beb..dc3b3727 100644
--- a/genno/core/computer.py
+++ b/genno/core/computer.py
@@ -108,9 +108,9 @@ def configure(
 
         Parameters
         ----------
-        path : .Path, *optional*
+        path : pathlib.Path, optional
             Path to a configuration file in JSON or YAML format.
-        fail : "raise" or str or :mod:`logging` level, *optional*
+        fail : "raise" or str or :mod:`logging` level, optional
             Passed to :meth:`.add_queue`. If not "raise", then log messages are
             generated for config handlers that fail. The Computer may be only partially
             configured.
@@ -149,7 +149,7 @@ def get_operator(self, name) -> Optional[Callable]:
 
         Returns
         -------
-        .callable
+        callable
         None
             If there is no callable with the given `name` in any of :attr:`modules`.
         """
@@ -251,7 +251,7 @@ def add(self, data, *args, **kwargs) -> Union[KeyLike, Tuple[KeyLike, ...]]:
 
         Returns
         -------
-        |KeyLike| or tuple of |KeyLike|
+        KeyLike or tuple of KeyLike
             Some or all of the keys added to the Computer.
 
         See also
@@ -342,12 +342,13 @@ def add_queue(
 
         Parameters
         ----------
-        queue : iterable of 2- or N-:class:`tuple`
-            The members of each tuple are the arguments (:class:`tuple`) and,
-            optionally, keyword arguments (e.g :class:`dict`) to :meth:`add`.
-        max_tries : int, *optional*
+        queue : iterable of tuple
+            Each item is either a N-:class:`tuple` of positional arguments to
+            :meth:`add`, or a 2-:class:`tuple` of (:class:`.tuple` of positional
+            arguments, :class:`dict` of keyword arguments).
+        max_tries : int, optional
             Retry adding elements up to this many times.
-        fail : "raise" or str or :mod:`logging` level, *optional*
+        fail : "raise" or str or :mod:`logging` level, optional
             Action to take when a computation from `queue` cannot be added after
             `max_tries`: "raise" an exception, or log messages on the indicated level
             and continue.
@@ -434,10 +435,10 @@ def add_single(
             A string, Key, or other value identifying the output of `computation`.
         computation : object
             Any computation. See :attr:`graph`.
-        strict : bool, *optional*
+        strict : bool, optional
             If True, `key` must not already exist in the Computer, and any keys
             referred to by `computation` must exist.
-        index : bool, *optional*
+        index : bool, optional
             If True, `key` is added to the index as a full-resolution key, so it can be
             later retrieved with :meth:`full_key`.
 
@@ -503,7 +504,7 @@ def apply(
 
         Parameters
         ----------
-        generator : .callable
+        generator : callable
             Function to apply to `keys`. This function **may** take a first positional
             argument annotated with :class:`.Computer` or a subtype; if so, then it is
             provided with a reference to `self`.
@@ -608,7 +609,7 @@ def get(self, key=None):
 
         Parameters
         ----------
-        key : str, *optional*
+        key : str, optional
             If not provided, :attr:`default_key` is used.
 
         Raises
@@ -646,7 +647,7 @@ def get(self, key=None):
     # Convenience methods for the graph and its keys
 
     def keys(self):
-        """Return the keys of :attr:`graph`."""
+        """Return the keys of :attr:`~genno.Computer.graph`."""
         return self.graph.keys()
 
     def full_key(self, name_or_key: KeyLike) -> KeyLike:
@@ -676,16 +677,16 @@ def check_keys(
 
         Parameters
         ----------
-        keys : |KeyLike|
+        keys : KeyLike
             Some :class:`Keys <Key>` or strings.
-        predicate : callable, *optional*
+        predicate : callable, optional
             Function to run on each of `keys`; see below.
-        action : "raise" or any other value
+        action : "raise" or str
             Action to take on missing `keys`.
 
         Returns
         -------
-        list of |KeyLike|
+        list of KeyLike
             One item for each item ``k`` in `keys`:
 
             1. ``k`` itself, unchanged, if `predicate` is given and ``predicate(k)``
@@ -748,16 +749,16 @@ def infer_keys(
 
         Parameters
         ----------
-        key_or_keys : |KeyLike| or list of |KeyLike|
-        dims : list of str, *optional*
+        key_or_keys : KeyLike or list of KeyLike
+        dims : list of str, optional
             Drop all but these dimensions from the returned key(s).
 
         Returns
         -------
-        |KeyLike|
-            If `key_or_keys` is a single |KeyLike|.
-        list of |KeyLike|
-            If `key_or_keys` is an iterable of |KeyLike|.
+        KeyLike
+            If `key_or_keys` is a single KeyLike.
+        list of KeyLike
+            If `key_or_keys` is an iterable of KeyLike.
         """
         single = isinstance(key_or_keys, (Key, Hashable))
         keys = [key_or_keys] if single else tuple(cast(Iterable, key_or_keys))
@@ -852,7 +853,7 @@ def add_product(self, *args, **kwargs):
         """Deprecated.
 
         .. deprecated:: 1.18.0
-           Instead use :func:`.add_mul` via:
+           Instead use :func:`.add_binop` via:
 
            .. code-block:: python
 
@@ -902,13 +903,13 @@ def aggregate(
             quantity.
         dims_or_groups: str or iterable of str or dict
             Name(s) of the dimension(s) to sum over, or nested dict.
-        weights : :class:`xarray.DataArray`, *optional*
+        weights : :class:`xarray.DataArray`, optional
             Weights for weighted aggregation.
-        keep : bool, *optional*
+        keep : bool, optional
             Passed to :meth:`operator.aggregate <genno.operator.aggregate>`.
-        sums : bool, *optional*
+        sums : bool, optional
             Passed to :meth:`add`.
-        fail : str or int, *optional*
+        fail : str or int, optional
             Passed to :meth:`add_queue` via :meth:`add`.
 
         Returns
diff --git a/genno/core/describe.py b/genno/core/describe.py
index d1a27182..8b7e234f 100644
--- a/genno/core/describe.py
+++ b/genno/core/describe.py
@@ -83,7 +83,8 @@ def label(arg, max_length=MAX_ITEM_LENGTH) -> str:
     The label depends on the type of `arg`:
 
     - :class:`.xarray.DataArray`: the first line of the string representation.
-    - :func:`.partial` object: a less-verbose version that omits None arguments.
+    - :func:`functools.partial` object: a less-verbose version that omits None
+      arguments.
     - Item protected with :func:`.dask.core.quote`: its literal value.
     - A callable, e.g. a function: its name.
     - Anything else: its :class:`str` representation.
diff --git a/genno/core/graph.py b/genno/core/graph.py
index f867b77d..d453e7e8 100644
--- a/genno/core/graph.py
+++ b/genno/core/graph.py
@@ -80,12 +80,14 @@ def __contains__(self, item) -> bool:
             return False
 
     def pop(self, *args):
+        """Overload :meth:`dict.pop` to also call :meth:`_deindex`."""
         try:
             return super().pop(*args)
         finally:
             self._deindex(args[0])
 
     def update(self, arg=None, **kwargs):
+        """Overload :meth:`dict.pop` to also call :meth:`_index`."""
         if isinstance(arg, (Sequence, Generator)):
             arg0, arg1 = tee(arg)
             arg_keys = map(itemgetter(0), arg0)
@@ -114,14 +116,14 @@ def infer(
 
         Parameters
         ----------
-        dims : list of str, *optional*
+        dims : list of str, optional
             Drop all but these dimensions from the returned key(s).
 
         Returns
         -------
         str
             If `key` is not found in the Graph.
-        Key
+        .Key
             `key` with either its full dimensions (cf. :meth:`full_key`) or, if `dims`
             are given, with only these dims.
         """
diff --git a/genno/core/key.py b/genno/core/key.py
index ff8c1be5..e384cfbb 100644
--- a/genno/core/key.py
+++ b/genno/core/key.py
@@ -113,22 +113,6 @@ def from_str_or_key(
     ) -> "Key":
         """Return a new Key from *value*.
 
-        Parameters
-        ----------
-        value : str or Key
-            Value to use to generate a new Key.
-        drop : list of str or :obj:`True`, *optional*
-            Existing dimensions of *value* to drop. See :meth:`drop`.
-        append : list of str, *optional*.
-            New dimensions to append to the returned Key. See :meth:`append`.
-        tag : str, *optional*
-            Tag for returned Key. If *value* has a tag, the two are joined
-            using a '+' character. See :meth:`add_tag`.
-
-        Returns
-        -------
-        :class:`Key`
-
         .. versionchanged:: 1.18.0
 
            Calling :meth:`from_str_or_key` with a single argument is no longer
@@ -141,6 +125,22 @@ def from_str_or_key(
 
               k1 = Key("foo:a-b-c:t1", drop="b", append="d", tag="t2")
               k2 = Key("foo:a-b-c:t1").drop("b").append("d)"
+
+        Parameters
+        ----------
+        value : str or .Key
+            Value to use to generate a new Key.
+        drop : list of str or :obj:`True`, optional
+            Existing dimensions of *value* to drop. See :meth:`drop`.
+        append : list of str, optional
+            New dimensions to append to the returned Key. See :meth:`append`.
+        tag : str, optional
+            Tag for returned Key. If *value* has a tag, the two are joined
+            using a '+' character. See :meth:`add_tag`.
+
+        Returns
+        -------
+        :class:`Key`
         """
         base = cls(value)
 
diff --git a/genno/core/operator.py b/genno/core/operator.py
index 63d9f656..f999adf0 100644
--- a/genno/core/operator.py
+++ b/genno/core/operator.py
@@ -1,11 +1,10 @@
 from functools import update_wrapper
 from inspect import signature
-from typing import TYPE_CHECKING, Any, Callable, ClassVar, Dict, Optional, Tuple, Union
+from typing import Any, Callable, ClassVar, Dict, Optional, Tuple, Union
 from warnings import warn
 
-if TYPE_CHECKING:
-    from .computer import Computer
-    from .key import KeyLike
+from .computer import Computer
+from .key import KeyLike
 
 
 class Operator:
@@ -70,7 +69,7 @@ def define(
 
         Parameters
         ----------
-        helper : Callable, *optional*
+        helper : Callable, optional
             Equivalent to calling :meth:`helper` on the Operator instance.
         """
 
@@ -111,13 +110,13 @@ def decorator(func: Callable) -> "Operator":
         return decorator
 
     def helper(
-        self, func: Callable[..., Union["KeyLike", Tuple["KeyLike", ...]]]
+        self, func: Callable[..., Union[KeyLike, Tuple[KeyLike, ...]]]
     ) -> Callable:
         """Register `func` as the convenience method for adding task(s)."""
         self.__class__._add_tasks = staticmethod(func)
         return func
 
-    def add_tasks(self, c: "Computer", *args, **kwargs) -> Tuple["KeyLike", ...]:
+    def add_tasks(self, c: "Computer", *args, **kwargs) -> Tuple[KeyLike, ...]:
         """Invoke :attr:`_add_task` to add tasks to `c`."""
         if self._add_tasks is None:
             raise NotImplementedError
diff --git a/genno/core/quantity.py b/genno/core/quantity.py
index fba183c6..eab65788 100644
--- a/genno/core/quantity.py
+++ b/genno/core/quantity.py
@@ -16,7 +16,7 @@
 class Quantity(DataArrayLike["Quantity"]):
     """A sparse data structure that behaves like :class:`xarray.DataArray`.
 
-    Depending on the value of :data:`CLASS`, Quantity is either :class:`.AttrSeries` or
+    Depending on the value of :data:`.CLASS`, Quantity is either :class:`.AttrSeries` or
     :class:`.SparseDataArray`.
     """
 
diff --git a/genno/core/sparsedataarray.py b/genno/core/sparsedataarray.py
index 0050dd44..24658dab 100644
--- a/genno/core/sparsedataarray.py
+++ b/genno/core/sparsedataarray.py
@@ -82,7 +82,7 @@ def COO_data(self):
 
     @property
     def dense(self):
-        """Return a copy with dense (:class:`.ndarray`) data."""
+        """Return a copy with dense (:class:`numpy.ndarray`) data."""
         try:
             # Use existing method xr.Variable._to_dense()
             return self.da._replace(variable=self.da.variable._to_dense())
@@ -92,7 +92,7 @@ def dense(self):
 
     @property
     def dense_super(self):
-        """Return a proxy to a :class:`.ndarray`-backed :class:`.DataArray`."""
+        """Return a proxy to a :class:`numpy.ndarray`-backed :class:`xarray.DataArray`."""
         return super(SparseDataArray, self.dense)
 
 
@@ -118,10 +118,11 @@ class SparseDataArray(OverrideItem, xr.DataArray, Quantity):
     """:class:`~xarray.DataArray` with sparse data.
 
     SparseDataArray uses :class:`sparse.COO` for storage with :data:`numpy.nan`
-    as its :attr:`sparse.COO.fill_value`. Some methods of :class:`~xarray.DataArray`
-    are overridden to ensure data is in sparse, or dense, format as necessary, to
-    provide expected functionality not currently supported by :mod:`sparse`, and to
-    avoid exhausting memory for some operations that require dense data.
+    as its :attr:`sparse.SparseArray.fill_value`. Some methods of
+    :class:`~xarray.DataArray` are overridden to ensure data is in sparse, or dense,
+    format as necessary, to provide expected functionality not currently supported by
+    :mod:`sparse`, and to avoid exhausting memory for some operations that require dense
+    data.
     """
 
     __slots__: Tuple[str, ...] = tuple()
@@ -245,7 +246,7 @@ def to_dataframe(
         name: Optional[Hashable] = None,
         dim_order: Optional[Sequence[Hashable]] = None,
     ) -> pd.DataFrame:
-        """Convert this array and its coords into a :class:`~xarray.DataFrame`.
+        """Convert this array and its coords into a :class:`pandas.DataFrame`.
 
         Overrides :meth:`~xarray.DataArray.to_dataframe`.
         """
diff --git a/genno/operator.py b/genno/operator.py
index 2a9ae8f6..5cdb0974 100644
--- a/genno/operator.py
+++ b/genno/operator.py
@@ -95,15 +95,15 @@ def add_binop(func, c: "Computer", key, *quantities, **kwargs) -> Key:
 
     Parameters
     ----------
-    key : str or Key
+    key : str or .Key
         Key or name of the new quantity. If a Key, any dimensions are ignored; the
         dimensions of the result are the union of the dimensions of `quantities`.
-    sums : bool, *optional*
+    sums : bool, optional
         If :obj:`True`, all partial sums of the new quantity are also added.
 
     Returns
     -------
-    Key
+    .Key
         The full key of the new quantity.
 
     Example
@@ -146,7 +146,7 @@ def add(*quantities: Quantity, fill_value: float = 0.0) -> Quantity:
 
     Returns
     -------
-    Quantity
+    .Quantity
         Units are the same as the first of `quantities`.
 
     See also
@@ -329,8 +329,8 @@ def broadcast_map(
     """Broadcast `quantity` using a `map`.
 
     The `map` must be a 2-dimensional Quantity with dimensions (``d1``, ``d2``), such as
-    returned by :func:`map_as_qty`. `quantity` must also have a dimension ``d1``.
-    Typically ``len(d2) > len(d1)``.
+    returned by :func:`ixmp.report.operator.map_as_qty`. `quantity` must also have a
+    dimension ``d1``. Typically ``len(d2) > len(d1)``.
 
     `quantity` is 'broadcast' by multiplying it with `map`, and then summing on the
     common dimension ``d1``. The result has the dimensions of `quantity`, but with
@@ -338,9 +338,10 @@ def broadcast_map(
 
     Parameters
     ----------
-    rename : dict (str -> str), *optional*
-        Dimensions to rename on the result.
-    strict : bool, *optional*
+    rename : dict, optional
+        Dimensions to rename on the result; mapping from original dimension
+        (:class:`str`) to target name (:class:`str`).
+    strict : bool, optional
         Require that each element of ``d2`` is mapped from exactly 1 element of ``d1``.
     """
     if strict and int(map.sum().item()) != len(map.coords[map.dims[1]]):
@@ -354,11 +355,11 @@ def combine(
     select: Optional[List[Mapping]] = None,
     weights: Optional[List[float]] = None,
 ) -> Quantity:  # noqa: F811
-    """Sum distinct *quantities* by *weights*.
+    """Sum distinct `quantities` by `weights`.
 
     Parameters
     ----------
-    *quantities : Quantity
+    *quantities : .Quantity
         The quantities to be added.
     select : list of dict
         Elements to be selected from each quantity. Must have the same number of
@@ -370,7 +371,7 @@ def combine(
     Raises
     ------
     ValueError
-        If the *quantities* have mismatched units.
+        If the `quantities` have mismatched units.
     """
     # Handle arguments
     if select is None:
@@ -495,8 +496,8 @@ def div(numerator: Union[Quantity, float], denominator: Quantity) -> Quantity:
 
     Parameters
     ----------
-    numerator : Quantity
-    denominator : Quantity
+    numerator : .Quantity
+    denominator : .Quantity
 
     See also
     --------
@@ -505,7 +506,7 @@ def div(numerator: Union[Quantity, float], denominator: Quantity) -> Quantity:
     return numerator / denominator
 
 
-#: Alias of :func:`div`, for backwards compatibility.
+#: Alias of :func:`~genno.operator.div`, for backwards compatibility.
 #:
 #: .. note:: This may be deprecated and possibly removed in a future version.
 ratio = div
@@ -605,7 +606,7 @@ def load_file(
     units: Optional[UnitLike] = None,
     name: Optional[str] = None,
 ) -> Any:
-    """Read the file at `path` and return its contents as a :class:`.Quantity`.
+    """Read the file at `path` and return its contents as a :class:`~genno.Quantity`.
 
     Some file formats are automatically converted into objects for direct use in genno
     computations:
@@ -624,7 +625,7 @@ def load_file(
     ----------
     path : pathlib.Path
         Path to the file to read.
-    dims : collections.abc.Collection or collections.abc.Mapping, *optional*
+    dims : collections.abc.Collection or collections.abc.Mapping, optional
         If a collection of names, other columns besides these and 'value' are discarded.
         If a mapping, the keys are the column labels in `path`, and the values are the
         target dimension names.
@@ -660,7 +661,7 @@ def add_load_file(func, c: "Computer", path, key=None, **kwargs):
     ----------
     path : os.PathLike
         Path to the file, e.g. '/path/to/foo.ext'.
-    key : str or Key, *optional*
+    key : str or .Key, optional
         Key for the quantity read from the file.
 
     Other parameters
@@ -767,7 +768,7 @@ def mul(*quantities: Quantity) -> Quantity:
     return reduce(operator.mul, quantities)
 
 
-#: Alias of :func:`mul`, for backwards compatibility.
+#: Alias of :func:`~genno.operator.mul`, for backwards compatibility.
 #:
 #: .. note:: This may be deprecated and possibly removed in a future version.
 product = mul
@@ -778,7 +779,7 @@ def pow(a: Quantity, b: Union[Quantity, int]) -> Quantity:
 
     Returns
     -------
-    Quantity
+    .Quantity
         If `b` is :class:`int` or a Quantity with all :class:`int` values that are equal
         to one another, then the quantity has the units of `a` raised to this power;
         for example, "kg²" → "kg⁴" if `b` is 2. In other cases, there are no meaningful
@@ -894,13 +895,17 @@ def select(
 
     Parameters
     ----------
-    indexers : dict (str -> xarray.DataArray or list of str)
-        Elements to be selected from `qty`. Mapping from dimension names to coords along
-        the respective dimension of `qty`, or to xarray-style indexers. Values not
-        appearing in the dimension coords are silently ignored.
-    inverse : bool, *optional*
+    indexers : dict
+        Elements to be selected from `qty`. Mapping from dimension names (:class:`str`)
+        to either:
+
+        - :class:`list` of `str`: coords along the respective dimension of `qty`, or
+        - :class:`xarray.DataArray`: xarray-style indexers.
+
+        Values not appearing in the dimension coords are silently ignored.
+    inverse : bool, optional
         If :obj:`True`, *remove* the items in indexers instead of keeping them.
-    drop : bool, *optional*
+    drop : bool, optional
         If :obj:`True`, drop dimensions that are indexed by a scalar value (for
         instance, :py:`"foo"` or :py:`999`) in `indexers`. Note that dimensions indexed
         by a length-1 list of labels (for instance :py:`["foo"]`) are not dropped; this
@@ -958,10 +963,10 @@ def sum(
 
     Parameters
     ----------
-    weights : .Quantity, *optional*
+    weights : .Quantity, optional
         If `dimensions` is given, `weights` must have at least these dimensions.
         Otherwise, any dimensions are valid.
-    dimensions : list of str, *optional*
+    dimensions : list of str, optional
         If not provided, sum over all dimensions. If provided, sum over these
         dimensions.
     """
diff --git a/genno/testing/__init__.py b/genno/testing/__init__.py
index 553effe1..7e015c5c 100644
--- a/genno/testing/__init__.py
+++ b/genno/testing/__init__.py
@@ -60,7 +60,7 @@ def add_large_data(c: Computer, num_params, N_dims=6, N_data=0):
     The result is a matrix wherein the Cartesian product of all the keys is very large—
     about 2e17 elements for N_dim = 6—but the contents are very sparse. This can be
     handled by :class:`.SparseDataArray`, but not by :class:`xarray.DataArray` backed
-    by :class:`np.array`.
+    by :class:`numpy.ndarray`.
     """
 
     def _fib():
@@ -234,7 +234,7 @@ def assert_logs(caplog, message_or_messages=None, at_level=None):
         The pytest caplog fixture.
     message_or_messages : str or list of str
         String(s) that must appear in log messages.
-    at_level : int, *optional*
+    at_level : int, optional
         Messages must appear on 'genno' or a sub-logger with at least this level.
     """
     __tracebackhide__ = True
@@ -289,12 +289,12 @@ def assert_qty_equal(
 
     Parameters
     ----------
-    check_type : bool, *optional*
+    check_type : bool, optional
         Assert that `a` and `b` are both :class:`.Quantity` instances. If :obj:`False`,
         the arguments are converted to Quantity.
-    check_attrs : bool, *optional*
+    check_attrs : bool, optional
         Also assert that check that attributes are identical.
-    ignore_extra_coords : bool, *optional*
+    ignore_extra_coords : bool, optional
         Ignore extra coords that are not dimensions. Only meaningful when Quantity is
         :class:`.SparseDataArray`.
     """
@@ -345,12 +345,12 @@ def assert_qty_allclose(
 
     Parameters
     ----------
-    check_type : bool, *optional*
+    check_type : bool, optional
         Assert that `a` and `b` are both :class:`.Quantity` instances. If :obj:`False`,
         the arguments are converted to Quantity.
-    check_attrs : bool, *optional*
+    check_attrs : bool, optional
         Also assert that check that attributes are identical.
-    ignore_extra_coords : bool, *optional*
+    ignore_extra_coords : bool, optional
         Ignore extra coords that are not dimensions. Only meaningful when Quantity is
         :class:`.SparseDataArray`.
     """
@@ -392,19 +392,20 @@ def assert_units(qty: Quantity, exp: str) -> None:
     ).dimensionless, f"Units '{qty.units:~}'; expected {repr(exp)}"
 
 
-def random_qty(shape: Dict[str, int], **kwargs):
+def random_qty(shape: Dict[str, int], **kwargs) -> Quantity:
     """Return a Quantity with `shape` and random contents.
 
     Parameters
     ----------
-    shape : dict (str -> int)
-        Mapping from dimension names to lengths along each dimension.
+    shape : dict
+        Mapping from dimension names (:class:`str`) to lengths along each dimension
+        (:class:`int`).
     **kwargs
-        Other keyword arguments to :class:`Quantity`.
+        Other keyword arguments to :class:`.Quantity`.
 
     Returns
     -------
-    Quantity
+    .Quantity
         Random data with one dimension for each key in `shape`, and coords along those
         dimensions like "foo1", "foo2", with total length matching the value from
         `shape`. If `shape` is empty, a scalar (0-dimensional) Quantity.
diff --git a/genno/testing/jupyter.py b/genno/testing/jupyter.py
index e2f8bd13..0819065c 100644
--- a/genno/testing/jupyter.py
+++ b/genno/testing/jupyter.py
@@ -16,11 +16,11 @@ def run_notebook(nb_path, tmp_path, env=None, **kwargs):
 
     Parameters
     ----------
-    nb_path : path-like
+    nb_path : os.PathLike
         The notebook file to execute.
-    tmp_path : path-like
+    tmp_path : os.PathLike
         A directory in which to create temporary output.
-    env : dict-like, *optional*
+    env : mapping, optional
         Execution environment for :mod:`nbclient`. Default: :obj:`os.environ`.
     kwargs :
         Keyword arguments for :class:`nbclient.NotebookClient`. Defaults are set for:
@@ -102,7 +102,7 @@ def get_cell_output(nb, name_or_index, kind="data"):
 
     Parameters
     ----------
-    kind : str, *optional*
+    kind : str, optional
         Kind of cell output to retrieve. For 'data', the data in format 'text/plain' is
         run through :func:`eval`. To retrieve an exception message, use 'evalue'.
     """
diff --git a/genno/util.py b/genno/util.py
index d918aabb..57e8879c 100644
--- a/genno/util.py
+++ b/genno/util.py
@@ -23,7 +23,7 @@
 #: - The '%' symbol cannot be supported by pint, because it is a Python operator; it is
 #:   replaced with “percent”.
 #:
-#: Additional values can be added with :func:`configure`; see :ref:`config-units`.
+#: Additional values can be added with :func:`.configure`; see :ref:`config-units`.
 REPLACE_UNITS = {
     "%": "percent",
 }
@@ -176,7 +176,7 @@ def free_parameters(func: Callable) -> Mapping:
     """Retrieve information on the free parameters of `func`.
 
     Identical to :py:`inspect.signature(func).parameters`; that is, to
-    :attr:`inspect.Signature.parameters`. :func:`free_pars` also:
+    :attr:`inspect.Signature.parameters`. :py:`free_parameters` also:
 
     - Handles functions that have been :func:`functools.partial`'d, returning only the
       parameters that have *not* already been assigned a value by the