brainstate 0.1.8__py2.py3-none-any.whl → 0.1.9__py2.py3-none-any.whl

brainstate/nn/_dynamics.py

@@ -1,1343 +1,1343 @@
-# Copyright 2024 BDP Ecosystem Limited. All Rights Reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-# ==============================================================================
-
-# -*- coding: utf-8 -*-
-
-
-"""
-All the basic dynamics class for the ``brainstate``.
-
-For handling dynamical systems:
-
-- ``DynamicsGroup``: The class for a group of modules, which update ``Projection`` first,
-                   then ``Dynamics``, finally others.
-- ``Projection``: The class for the synaptic projection.
-- ``Dynamics``: The class for the dynamical system.
-
-For handling the delays:
-
-- ``Delay``: The class for all delays.
-- ``DelayAccess``: The class for the delay access.
-
-"""
-
-from typing import Any, Dict, Callable, Hashable, Optional, Union, TypeVar, TYPE_CHECKING, Tuple
-
-import jax
-import numpy as np
-
-from brainstate import environ
-from brainstate._state import State
-from brainstate.graph import Node
-from brainstate.mixin import ParamDescriber
-from brainstate.typing import Size, ArrayLike
-from ._delay import StateWithDelay, Delay
-from ._module import Module
-
-__all__ = [
-    'DynamicsGroup', 'Projection', 'Dynamics',
-    'Prefetch', 'PrefetchDelay', 'PrefetchDelayAt', 'OutputDelayAt',
-]
-
-T = TypeVar('T')
-_max_order = 10
-
-
-class Projection(Module):
-    """
-    Base class for synaptic projection modules in neural network modeling.
-
-    This class defines the interface for modules that handle projections between
-    neural populations. Projections process input signals and transform them
-    before they reach the target neurons, implementing the connectivity patterns
-    in neural networks.
-
-    In the BrainState execution order, Projection modules are updated before
-    Dynamics modules, following the natural information flow in neural systems:
-    1. Projections process inputs (synaptic transmission)
-    2. Dynamics update neuron states (neural integration)
-
-    The Projection class does not implement the update logic directly but delegates
-    to its child nodes. If no child nodes exist, it raises a ValueError.
-
-    Parameters
-    ----------
-    *args : Any
-        Arguments passed to the parent Module class.
-    **kwargs : Any
-        Keyword arguments passed to the parent Module class.
-
-    Raises
-    ------
-    ValueError
-        If the update() method is called but no child nodes are defined.
-
-    Notes
-    -----
-    Derived classes should implement specific projection behaviors, such as
-    dense connectivity, sparse connectivity, or specific weight update rules.
-    """
-    __module__ = 'brainstate.nn'
-
-    def update(self, *args, **kwargs):
-        sub_nodes = tuple(self.nodes(allowed_hierarchy=(1, 1)).values())
-        if len(sub_nodes):
-            for node in sub_nodes:
-                node(*args, **kwargs)
-        else:
-            raise ValueError('Do not implement the update() function.')
-
-
-class Dynamics(Module):
-    """
-    Base class for implementing neural dynamics models in BrainState.
-
-    Dynamics classes represent the core computational units in neural simulations,
-    implementing the differential equations or update rules that govern neural activity.
-    This class provides infrastructure for managing neural populations, handling inputs,
-    and coordinating updates within the simulation framework.
-
-    The Dynamics class serves several key purposes:
-    1. Managing neuron population geometry and size information
-    2. Handling current and delta (instantaneous change) inputs to neurons
-    3. Supporting before/after update hooks for computational dependencies
-    4. Providing access to delayed state variables through the prefetch mechanism
-    5. Establishing the execution order in neural network simulations
-
-    Parameters
-    ----------
-    in_size : Size
-        The geometry of the neuron population. Can be an integer (e.g., 10) for
-        1D neuron arrays, or a tuple (e.g., (10, 10)) for multi-dimensional populations.
-    name : Optional[str], default=None
-        Optional name identifier for this dynamics module.
-
-    Attributes
-    ----------
-    in_size : tuple
-        The shape/geometry of the neuron population.
-    out_size : tuple
-        The output shape, typically matches in_size.
-    current_inputs : Optional[Dict[str, Union[Callable, ArrayLike]]]
-        Dictionary of registered current input functions or arrays.
-    delta_inputs : Optional[Dict[str, Union[Callable, ArrayLike]]]
-        Dictionary of registered delta input functions or arrays.
-    before_updates : Optional[Dict[Hashable, Callable]]
-        Dictionary of functions to call before the main update.
-    after_updates : Optional[Dict[Hashable, Callable]]
-        Dictionary of functions to call after the main update.
-
-    Notes
-    -----
-    In the BrainState execution sequence, Dynamics modules are updated after
-    Projection modules and before other module types, reflecting the natural
-    flow of information in neural systems.
-
-    There are several essential attributes:
-
-    - ``size``: the geometry of the neuron group. For example, `(10, )` denotes a line of
-      neurons, `(10, 10)` denotes a neuron group aligned in a 2D space, `(10, 15, 4)` denotes
-      a 3-dimensional neuron group.
-    - ``num``: the flattened number of neurons in the group. For example, `size=(10, )` => \
-      `num=10`, `size=(10, 10)` => `num=100`, `size=(10, 15, 4)` => `num=600`.
-
-
-    See Also
-    --------
-    Module : Parent class providing base module functionality
-    Projection : Class for handling synaptic projections between neural populations
-    DynamicsGroup : Container for organizing multiple dynamics modules
-    """
-
-    __module__ = 'brainstate.nn'
-
-    graph_invisible_attrs = ('_before_updates', '_after_updates', '_current_inputs', '_delta_inputs')
-
-    # before updates
-    _before_updates: Optional[Dict[Hashable, Callable]]
-
-    # after updates
-    _after_updates: Optional[Dict[Hashable, Callable]]
-
-    # current inputs
-    _current_inputs: Optional[Dict[str, ArrayLike | Callable]]
-
-    # delta inputs
-    _delta_inputs: Optional[Dict[str, ArrayLike | Callable]]
-
-    def __init__(
-        self,
-        in_size: Size,
-        name: Optional[str] = None,
-    ):
-        # initialize
-        super().__init__(name=name)
-
-        # geometry size of neuron population
-        if isinstance(in_size, (list, tuple)):
-            if len(in_size) <= 0:
-                raise ValueError(f'"in_size" must be int, or a tuple/list of int. But we got {type(in_size)}')
-            if not isinstance(in_size[0], (int, np.integer)):
-                raise ValueError(f'"in_size" must be int, or a tuple/list of int. But we got {type(in_size)}')
-            in_size = tuple(in_size)
-        elif isinstance(in_size, (int, np.integer)):
-            in_size = (in_size,)
-        else:
-            raise ValueError(f'"in_size" must be int, or a tuple/list of int. But we got {type(in_size)}')
-        self.in_size = in_size
-
-        # current inputs
-        self._current_inputs = None
-
-        # delta inputs
-        self._delta_inputs = None
-
-        # before updates
-        self._before_updates = None
-
-        # after updates
-        self._after_updates = None
-
-        # in-/out- size of neuron population
-        self.out_size = self.in_size
-
-    # def __pretty_repr_item__(self, name, value):
-    #     if name in [
-    #         '_before_updates', '_after_updates', '_current_inputs', '_delta_inputs',
-    #         '_in_size', '_out_size', '_name', '_mode',
-    #     ]:
-    #         return (name, value) if value is None else (name[1:], value)  # skip the first `_`
-    #     return super().__pretty_repr_item__(name, value)
-
-    @property
-    def varshape(self):
-        """
-        Get the shape of variables in the neuron group.
-
-        This property provides access to the geometry (shape) of the neuron population,
-        which determines how variables and states are structured.
-
-        Returns
-        -------
-        tuple
-            A tuple representing the dimensional shape of the neuron group,
-            matching the in_size parameter provided during initialization.
-
-        See Also
-        --------
-        in_size : The input geometry specification for the neuron group
-        """
-        return self.in_size
-
-    @property
-    def current_inputs(self):
-        """
-        Get the dictionary of current inputs registered with this dynamics model.
-
-        Current inputs represent direct input currents that flow into the model.
-
-        Returns
-        -------
-        dict or None
-            A dictionary mapping keys to current input functions or values,
-            or None if no current inputs have been registered.
-
-        See Also
-        --------
-        add_current_input : Register a new current input
-        sum_current_inputs : Apply and sum all current inputs
-        delta_inputs : Dictionary of instantaneous change inputs
-        """
-        return self._current_inputs
-
-    @property
-    def delta_inputs(self):
-        """
-        Get the dictionary of delta inputs registered with this dynamics model.
-
-        Delta inputs represent instantaneous changes to state variables (dX/dt).
-
-        Returns
-        -------
-        dict or None
-            A dictionary mapping keys to delta input functions or values,
-            or None if no delta inputs have been registered.
-
-        See Also
-        --------
-        add_delta_input : Register a new delta input
-        sum_delta_inputs : Apply and sum all delta inputs
-        current_inputs : Dictionary of direct current inputs
-        """
-        return self._delta_inputs
-
-    def add_current_input(
-        self,
-        key: str,
-        inp: Union[Callable, ArrayLike],
-        label: Optional[str] = None
-    ):
-        """
-        Add a current input function or array to the dynamics model.
-
-        Current inputs represent direct input currents that can be accessed during
-        model updates through the `sum_current_inputs()` method.
-
-        Parameters
-        ----------
-        key : str
-            Unique identifier for this current input. Used to retrieve or reference
-            the input later.
-        inp : Union[Callable, ArrayLike]
-            The input data or function that generates input data.
-            - If callable: Will be called during updates with arguments passed to `sum_current_inputs()`
-            - If array-like: Will be applied once and then automatically removed from available inputs
-        label : Optional[str], default=None
-            Optional grouping label for the input. When provided, allows selective
-            processing of inputs by label in `sum_current_inputs()`.
-
-        Raises
-        ------
-        ValueError
-            If the key has already been used for a different current input.
-
-        Notes
-        -----
-        - Inputs with the same label can be processed together using the `label`
-          parameter in `sum_current_inputs()`.
-        - Non-callable inputs are consumed when used (removed after first use).
-        - Callable inputs persist and can be called repeatedly.
-
-        See Also
-        --------
-        sum_current_inputs : Sum all current inputs matching a given label
-        add_delta_input : Add a delta input function or array
-        """
-        key = _input_label_repr(key, label)
-        if self._current_inputs is None:
-            self._current_inputs = dict()
-        if key in self._current_inputs:
-            if id(self._current_inputs[key]) != id(inp):
-                raise ValueError(f'Key "{key}" has been defined and used in the current inputs of {self}.')
-        self._current_inputs[key] = inp
-
-    def add_delta_input(
-        self,
-        key: str,
-        inp: Union[Callable, ArrayLike],
-        label: Optional[str] = None
-    ):
-        """
-        Add a delta input function or array to the dynamics model.
-
-        Delta inputs represent instantaneous changes to the model state (i.e., dX/dt contributions).
-        This method registers a function or array that provides delta inputs which will be
-        accessible during model updates through the `sum_delta_inputs()` method.
-
-        Parameters
-        ----------
-        key : str
-            Unique identifier for this delta input. Used to retrieve or reference
-            the input later.
-        inp : Union[Callable, ArrayLike]
-            The input data or function that generates input data.
-            - If callable: Will be called during updates with arguments passed to `sum_delta_inputs()`
-            - If array-like: Will be applied once and then automatically removed from available inputs
-        label : Optional[str], default=None
-            Optional grouping label for the input. When provided, allows selective
-            processing of inputs by label in `sum_delta_inputs()`.
-
-        Raises
-        ------
-        ValueError
-            If the key has already been used for a different delta input.
-
-        Notes
-        -----
-        - Inputs with the same label can be processed together using the `label`
-          parameter in `sum_delta_inputs()`.
-        - Non-callable inputs are consumed when used (removed after first use).
-        - Callable inputs persist and can be called repeatedly.
-
-        See Also
-        --------
-        sum_delta_inputs : Sum all delta inputs matching a given label
-        add_current_input : Add a current input function or array
-        """
-        key = _input_label_repr(key, label)
-        if self._delta_inputs is None:
-            self._delta_inputs = dict()
-        if key in self._delta_inputs:
-            if id(self._delta_inputs[key]) != id(inp):
-                raise ValueError(f'Key "{key}" has been defined and used.')
-        self._delta_inputs[key] = inp
-
-    def get_input(self, key: str):
-        """
-        Get a registered input function by its key.
-
-        Retrieves either a current input or a delta input function that was previously
-        registered with the given key. This method checks both current_inputs and
-        delta_inputs dictionaries for the specified key.
-
-        Parameters
-        ----------
-        key : str
-            The unique identifier used when the input function was registered.
-
-        Returns
-        -------
-        Callable or ArrayLike
-            The input function or array associated with the given key.
-
-        Raises
-        ------
-        ValueError
-            If no input function is found with the specified key in either
-            current_inputs or delta_inputs.
-
-        See Also
-        --------
-        add_current_input : Register a current input function
-        add_delta_input : Register a delta input function
-
-        Examples
-        --------
-        >>> model = Dynamics(10)
-        >>> model.add_current_input('stimulus', lambda t: np.sin(t))
-        >>> input_func = model.get_input('stimulus')
-        >>> input_func(0.5)  # Returns sin(0.5)
-        """
-        if self._current_inputs is not None and key in self._current_inputs:
-            return self._current_inputs[key]
-        elif self._delta_inputs is not None and key in self._delta_inputs:
-            return self._delta_inputs[key]
-        else:
-            raise ValueError(f'Input key {key} is not in current/delta inputs of the module {self}.')
-
-    def sum_current_inputs(
-        self,
-        init: Any,
-        *args,
-        label: Optional[str] = None,
-        **kwargs
-    ):
-        """
-        Summarize all current inputs by applying and summing all registered current input functions.
-
-        This method iterates through all registered current input functions (from `.current_inputs`)
-        and applies them to calculate the total input current for the dynamics model. It adds all results
-        to the initial value provided.
-
-        Parameters
-        ----------
-        init : Any
-            The initial value to which all current inputs will be added.
-        *args : tuple
-            Variable length argument list passed to each current input function.
-        label : Optional[str], default=None
-            If provided, only process current inputs with this label prefix.
-            When None, process all current inputs regardless of label.
-        **kwargs : dict
-            Arbitrary keyword arguments passed to each current input function.
-
-        Returns
-        -------
-        Any
-            The initial value plus all applicable current inputs summed together.
-
-        Notes
-        -----
-        - Non-callable current inputs are applied once and then automatically removed from
-          the current_inputs dictionary.
-        - Callable current inputs remain registered for subsequent calls.
-        - When a label is provided, only current inputs with keys starting with that label
-          are applied.
-        """
-        if self._current_inputs is None:
-            return init
-        if label is None:
-            filter_fn = lambda k: True
-        else:
-            label_repr = _input_label_start(label)
-            filter_fn = lambda k: k.startswith(label_repr)
-        for key in tuple(self._current_inputs.keys()):
-            if filter_fn(key):
-                out = self._current_inputs[key]
-                if callable(out):
-                    try:
-                        init = init + out(*args, **kwargs)
-                    except Exception as e:
-                        raise ValueError(
-                            f'Error in delta input value {key}: {out}\n'
-                            f'Error: {e}'
-                        ) from e
-                else:
-                    try:
-                        init = init + out
-                    except Exception as e:
-                        raise ValueError(
-                            f'Error in delta input value {key}: {out}\n'
-                            f'Error: {e}'
-                        ) from e
-                    self._current_inputs.pop(key)
-        return init
-
-    def sum_delta_inputs(
-        self,
-        init: Any,
-        *args,
-        label: Optional[str] = None,
-        **kwargs
-    ):
-        """
-        Summarize all delta inputs by applying and summing all registered delta input functions.
-
-        This method iterates through all registered delta input functions (from `.delta_inputs`)
-        and applies them to calculate instantaneous changes to model states. It adds all results
-        to the initial value provided.
-
-        Parameters
-        ----------
-        init : Any
-            The initial value to which all delta inputs will be added.
-        *args : tuple
-            Variable length argument list passed to each delta input function.
-        label : Optional[str], default=None
-            If provided, only process delta inputs with this label prefix.
-            When None, process all delta inputs regardless of label.
-        **kwargs : dict
-            Arbitrary keyword arguments passed to each delta input function.
-
-        Returns
-        -------
-        Any
-            The initial value plus all applicable delta inputs summed together.
-
-        Notes
-        -----
-        - Non-callable delta inputs are applied once and then automatically removed from
-          the delta_inputs dictionary.
-        - Callable delta inputs remain registered for subsequent calls.
-        - When a label is provided, only delta inputs with keys starting with that label
-          are applied.
-        """
-        if self._delta_inputs is None:
-            return init
-        if label is None:
-            filter_fn = lambda k: True
-        else:
-            label_repr = _input_label_start(label)
-            filter_fn = lambda k: k.startswith(label_repr)
-        for key in tuple(self._delta_inputs.keys()):
-            if filter_fn(key):
-                out = self._delta_inputs[key]
-                if callable(out):
-                    try:
-                        init = init + out(*args, **kwargs)
-                    except Exception as e:
-                        raise ValueError(
-                            f'Error in delta input function {key}: {out}\n'
-                            f'Error: {e}'
-                        ) from e
-                else:
-                    try:
-                        init = init + out
-                    except Exception as e:
-                        raise ValueError(
-                            f'Error in delta input value {key}: {out}\n'
-                            f'Error: {e}'
-                        ) from e
-                    self._delta_inputs.pop(key)
-        return init
-
-    @property
-    def before_updates(self):
-        """
-        Get the dictionary of functions to execute before the module's update.
-
-        Returns
-        -------
-        dict or None
-            Dictionary mapping keys to callable functions that will be executed
-            before the main update, or None if no before updates are registered.
-
-        Notes
-        -----
-        Before updates are executed in the order they were registered whenever
-        the module is called via __call__.
-        """
-        return self._before_updates
-
-    @property
-    def after_updates(self):
-        """
-        Get the dictionary of functions to execute after the module's update.
-
-        Returns
-        -------
-        dict or None
-            Dictionary mapping keys to callable functions that will be executed
-            after the main update, or None if no after updates are registered.
-
-        Notes
-        -----
-        After updates are executed in the order they were registered whenever
-        the module is called via __call__, and may optionally receive the return
-        value from the update method.
-        """
-        return self._after_updates
-
-    def _add_before_update(self, key: Any, fun: Callable):
-        """
-        Register a function to be executed before the module's update.
-
-        Parameters
-        ----------
-        key : Any
-            A unique identifier for the update function.
-        fun : Callable
-            The function to execute before the module's update.
-
-        Raises
-        ------
-        KeyError
-            If the key is already registered in before_updates.
-
-        Notes
-        -----
-        Internal method used by the module system to register dependencies.
-        """
-        if self._before_updates is None:
-            self._before_updates = dict()
-        if key in self.before_updates:
-            raise KeyError(f'{key} has been registered in before_updates of {self}')
-        self.before_updates[key] = fun
-
-    def _add_after_update(self, key: Any, fun: Callable):
-        """
-        Register a function to be executed after the module's update.
-
-        Parameters
-        ----------
-        key : Any
-            A unique identifier for the update function.
-        fun : Callable
-            The function to execute after the module's update.
-
-        Raises
-        ------
-        KeyError
-            If the key is already registered in after_updates.
-
-        Notes
-        -----
-        Internal method used by the module system to register dependencies.
-        """
-        if self._after_updates is None:
-            self._after_updates = dict()
-        if key in self.after_updates:
-            raise KeyError(f'{key} has been registered in after_updates of {self}')
-        self.after_updates[key] = fun
-
-    def _get_before_update(self, key: Any):
-        """
-        Retrieve a registered before-update function by its key.
-
-        Parameters
-        ----------
-        key : Any
-            The identifier of the before-update function to retrieve.
-
-        Returns
-        -------
-        Callable
-            The registered before-update function.
-
-        Raises
-        ------
-        KeyError
-            If the key is not registered in before_updates or if before_updates is None.
-        """
-        if self._before_updates is None:
-            raise KeyError(f'{key} is not registered in before_updates of {self}')
-        if key not in self.before_updates:
-            raise KeyError(f'{key} is not registered in before_updates of {self}')
-        return self.before_updates.get(key)
-
-    def _get_after_update(self, key: Any):
-        """
-        Retrieve a registered after-update function by its key.
-
-        Parameters
-        ----------
-        key : Any
-            The identifier of the after-update function to retrieve.
-
-        Returns
-        -------
-        Callable
-            The registered after-update function.
-
-        Raises
-        ------
-        KeyError
-            If the key is not registered in after_updates or if after_updates is None.
-        """
-        if self._after_updates is None:
-            raise KeyError(f'{key} is not registered in after_updates of {self}')
-        if key not in self.after_updates:
-            raise KeyError(f'{key} is not registered in after_updates of {self}')
-        return self.after_updates.get(key)
-
-    def _has_before_update(self, key: Any):
-        """
-        Check if a before-update function is registered with the given key.
-
-        Parameters
-        ----------
-        key : Any
-            The identifier to check for in the before_updates dictionary.
-
-        Returns
-        -------
-        bool
-            True if the key is registered in before_updates, False otherwise.
-        """
-        if self._before_updates is None:
-            return False
-        return key in self.before_updates
-
-    def _has_after_update(self, key: Any):
-        """
-        Check if an after-update function is registered with the given key.
-
-        Parameters
-        ----------
-        key : Any
-            The identifier to check for in the after_updates dictionary.
-
-        Returns
-        -------
-        bool
-            True if the key is registered in after_updates, False otherwise.
-        """
-        if self._after_updates is None:
-            return False
-        return key in self.after_updates
-
-    def __call__(self, *args, **kwargs):
-        """
-        The shortcut to call ``update`` methods.
-        """
-
-        # ``before_updates``
-        if self.before_updates is not None:
-            for model in self.before_updates.values():
-                if hasattr(model, '_receive_update_input'):
-                    model(*args, **kwargs)
-                else:
-                    model()
-
-        # update the model self
-        ret = self.update(*args, **kwargs)
-
-        # ``after_updates``
-        if self.after_updates is not None:
-            for model in self.after_updates.values():
-                if hasattr(model, '_not_receive_update_output'):
-                    model()
-                else:
-                    model(ret)
-        return ret
-
-    def prefetch(self, item: str) -> 'Prefetch':
-        """
-        Create a reference to a state or variable that may not be initialized yet.
-
-        This method allows accessing module attributes or states before they are
-        fully defined, acting as a placeholder that will be resolved when called.
-        Particularly useful for creating references to variables that will be defined
-        during initialization or runtime.
-
-        Parameters
-        ----------
-        item : str
-            The name of the attribute or state to reference.
-
-        Returns
-        -------
-        Prefetch
-            A Prefetch object that provides access to the referenced item.
-
-        Examples
-        --------
-        >>> import brainstate
-        >>> import brainunit as u
-        >>> neuron = brainstate.nn.LIF(...)
-        >>> v_ref = neuron.prefetch('V')  # Reference to voltage
-        >>> v_value = v_ref()  # Get current value
-        >>> delayed_v = v_ref.delay.at(5.0 * u.ms)  # Get delayed value
-        """
-        return Prefetch(self, item)
-
-    def align_pre(self, dyn: Union[ParamDescriber[T], T]) -> T:
-        """
-        Registers a dynamics module to execute after this module.
-
-        This method establishes a sequential execution relationship where the specified
-        dynamics module will be called after this module completes its update. This
-        creates a feed-forward connection in the computational graph.
-
-        Parameters
-        ----------
-        dyn : Union[ParamDescriber[T], T]
-            The dynamics module to be executed after this module. Can be either:
-            - An instance of Dynamics
-            - A ParamDescriber that can instantiate a Dynamics object
-
-        Returns
-        -------
-        T
-            The dynamics module that was registered, allowing for method chaining.
-
-        Raises
-        ------
-        TypeError
-            If the input is not a Dynamics instance or a ParamDescriber that creates
-            a Dynamics instance.
-
-        Examples
-        --------
-        >>> import brainstate
-        >>> n1 = brainstate.nn.LIF(10)
-        >>> n1.align_pre(brainstate.nn.Expon.desc(n1.varshape))  # n2 will run after n1
-        """
-        if isinstance(dyn, Dynamics):
-            self._add_after_update(id(dyn), dyn)
-            return dyn
-        elif isinstance(dyn, ParamDescriber):
-            if not issubclass(dyn.cls, Dynamics):
-                raise TypeError(f'The input {dyn} should be an instance of {Dynamics}.')
-            if not self._has_after_update(dyn.identifier):
-                self._add_after_update(
-                    dyn.identifier,
-                    dyn() if ('in_size' in dyn.kwargs or len(dyn.args) > 0) else dyn(in_size=self.varshape)
-                )
-            return self._get_after_update(dyn.identifier)
-        else:
-            raise TypeError(f'The input {dyn} should be an instance of {Dynamics} or a delayed initializer.')
-
-    def prefetch_delay(self, state: str, delay_time, init: Callable = None) -> 'PrefetchDelayAt':
-        """
-        Create a reference to a delayed state or variable in the module.
-
-        This method simplifies the process of accessing a delayed version of a state or variable
-        within the module. It first creates a prefetch reference to the specified state,
-        then specifies the delay time for accessing this state.
-
-        Args:
-            state (str): The name of the state or variable to reference.
-            delay_time (ArrayLike): The amount of time to delay the variable access,
-                typically in time units (e.g., milliseconds).
-            init (Callable, optional): An optional initialization function to provide
-                a default value if the delayed state is not yet available.
-
-        Returns:
-            PrefetchDelayAt: An object that provides access to the variable at the specified delay time.
-        """
-        return PrefetchDelayAt(self, state, delay_time, init=init)
-
-    def output_delay(self, *delay_time) -> 'OutputDelayAt':
-        """
-        Create a reference to the delayed output of the module.
-
-        This method simplifies the process of accessing a delayed version of the module's output.
-        It instantiates an `OutputDelayAt` object, which can be used to retrieve the output value
-        at the specified delay time.
-
-        Args:
-            delay (Optional[ArrayLike]): The amount of time to delay the output access,
-                typically in time units (e.g., milliseconds). Defaults to None.
-
-        Returns:
-            OutputDelayAt: An object that provides access to the module's output at the specified delay time.
-        """
-        return OutputDelayAt(self, delay_time)
-
-
-class Prefetch(Node):
-    """
-    Prefetch a state or variable in a module before it is initialized.
-
-
-    This class provides a mechanism to reference a module's state or attribute
-    that may not have been initialized yet. It acts as a placeholder or reference
-    that will be resolved when called.
-
-    Use cases:
-    - Access variables within dynamics modules that will be defined later
-    - Create references to states across module boundaries
-    - Enable access to delayed states through the `.delay` property
-
-    Parameters
-    ----------
-    module : Module
-        The module that contains or will contain the referenced item.
-    item : str
-        The attribute name of the state or variable to prefetch.
-
-    Examples
-    --------
-    >>> import brainstate
-    >>> import brainunit as u
-    >>> neuron = brainstate.nn.LIF(...)
-    >>> v_reference = neuron.prefetch('V')  # Reference to voltage before initialization
-    >>> v_value = v_reference()  # Get the current value
-    >>> delay_ref = v_reference.delay.at(5.0 * u.ms)  # Reference voltage delayed by 5ms
-
-    Notes
-    -----
-    When called, this class retrieves the current value of the referenced item.
-    Use the `.delay` property to access delayed versions of the state.
-
-    """
-
-    def __init__(self, module: Dynamics, item: str):
-        """
-        Initialize a Prefetch object.
-
-        Parameters
-        ----------
-        module : Module
-            The module that contains or will contain the referenced item.
-        item : str
-            The attribute name of the state or variable to prefetch.
-        """
-        super().__init__()
-        self.module = module
-        self.item = item
-
-    @property
-    def delay(self):
-        """
-        Access delayed versions of the prefetched item.
-
-        Returns
-        -------
-        PrefetchDelay
-            An object that provides access to delayed versions of the prefetched item.
-        """
-        return PrefetchDelay(self.module, self.item)
-        # return PrefetchDelayAt(self.module, self.item, time)
-
-    def __call__(self, *args, **kwargs):
-        """
-        Get the current value of the prefetched item.
-
-        Returns
-        -------
-        Any
-            The current value of the referenced item. If the item is a State object,
-            returns its value attribute, otherwise returns the item itself.
-        """
-        item = _get_prefetch_item(self)
-        return item.value if isinstance(item, State) else item
-
-    def get_item_value(self):
-        """
-        Get the current value of the prefetched item.
-
-        Similar to __call__, but explicitly named for clarity.
-
-        Returns
-        -------
-        Any
-            The current value of the referenced item. If the item is a State object,
-            returns its value attribute, otherwise returns the item itself.
-        """
-        item = _get_prefetch_item(self)
-        return item.value if isinstance(item, State) else item
-
-    def get_item(self):
-        """
-        Get the referenced item object itself, not its value.
-
-        Returns
-        -------
-        Any
-            The actual referenced item from the module, which could be a State
-            object or any other attribute.
-        """
-        return _get_prefetch_item(self)
-
-
-class PrefetchDelay(Node):
-    """
-    Provides access to delayed versions of a prefetched state or variable.
-
-    This class acts as an intermediary for accessing delayed values of module variables.
-    It doesn't retrieve values directly but provides methods to specify the delay time
-    via the `at()` method.
-
-    Parameters
-    ----------
-    module : Dynamics
-        The dynamics module that contains the referenced state or variable.
-    item : str
-        The name of the state or variable to access with delay.
-
-    Examples
-    --------
-    >>> import brainstate
-    >>> import brainunit as u
-    >>> neuron = brainstate.nn.LIF(10)
-    >>> # Access voltage delayed by 5ms
-    >>> delayed_v = neuron.prefetch('V').delay.at(5.0 * u.ms)
-    >>> delayed_value = delayed_v()  # Get the delayed value
-    """
-
-    def __init__(self, module: Dynamics, item: str):
-        self.module = module
-        self.item = item
-
-    def at(self, *delay_time):
-        """
-        Specifies the delay time for accessing the variable.
-
-        Parameters
-        ----------
-        time : ArrayLike
-            The amount of time to delay the variable access, typically in time units
-            (e.g., milliseconds).
-
-        Returns
-        -------
-        PrefetchDelayAt
-            An object that provides access to the variable at the specified delay time.
-        """
-        return PrefetchDelayAt(self.module, self.item, delay_time)
-
-
-class PrefetchDelayAt(Node):
-    """
-    Provides access to a specific delayed state or variable value at the specific time.
-
-    This class represents the final step in the prefetch delay chain, providing
-    actual access to state values at a specific delay time. It converts the
-    specified time delay into steps and registers the delay with the appropriate
-    StateWithDelay handler.
-
-    Parameters
-    ----------
-    module : Dynamics
-        The dynamics module that contains the referenced state or variable.
-    item : str
-        The name of the state or variable to access with delay.
-    time : ArrayLike
-        The amount of time to delay access by, typically in time units (e.g., milliseconds).
-
-    Examples
-    --------
-    >>> import brainstate
-    >>> import brainunit as u
-    >>> neuron = brainstate.nn.LIF(10)
-    >>> # Create a reference to voltage delayed by 5ms
-    >>> delayed_v = PrefetchDelayAt(neuron, 'V', 5.0 * u.ms)
-    >>> # Get the delayed value
-    >>> v_value = delayed_v()
-    """
-
-    def __init__(
-        self,
-        module: Dynamics,
-        item: str,
-        delay_time: Tuple,
-        init: Callable = None
-    ):
-        """
-        Initialize a PrefetchDelayAt object.
-
-        Parameters
-        ----------
-        module : Dynamics
-            The dynamics module that contains the referenced state or variable.
-        item : str
-            The name of the state or variable to access with delay.
-        delay_time : Tuple
-            The amount of time to delay access by, typically in time units (e.g., milliseconds).
-        """
-        super().__init__()
-        assert isinstance(module, Dynamics), 'The module should be an instance of Dynamics.'
-        self.module = module
-        self.item = item
-        if not isinstance(delay_time, (tuple, list)):
-            delay_time = (delay_time,)
-        self.delay_time = delay_time
-        if len(delay_time) > 0:
-            key = _get_prefetch_delay_key(item)
-            if not module._has_after_update(key):
-                module._add_after_update(
-                    key,
-                    not_receive_update_output(
-                        StateWithDelay(module, item, init=init)
-                    )
-                )
-            self.state_delay: StateWithDelay = module._get_after_update(key)
-            self.delay_info = self.state_delay.register_delay(*delay_time)
-
-    def __call__(self, *args, **kwargs):
-        """
-        Retrieve the value of the state at the specified delay time.
-
-        Returns
-        -------
-        Any
-            The value of the state or variable at the specified delay time.
-        """
-        if len(self.delay_time) == 0:
-            return _get_prefetch_item(self).value
-        else:
-            return self.state_delay.retrieve_at_step(*self.delay_info)
-
-
-class OutputDelayAt(Node):
-    """
-    Provides access to a specific delayed state or variable value at the specific time.
-
-    This class represents the final step in the prefetch delay chain, providing
-    actual access to state values at a specific delay time. It converts the
-    specified time delay into steps and registers the delay with the appropriate
-    StateWithDelay handler.
-
-    Parameters
-    ----------
-    module : Dynamics
-        The dynamics module that contains the referenced state or variable.
-    time : ArrayLike
-        The amount of time to delay access by, typically in time units (e.g., milliseconds).
-
-    Examples
-    --------
-    >>> import brainstate
-    >>> import brainunit as u
-    >>> neuron = brainstate.nn.LIF(10)
-    >>> # Create a reference to voltage delayed by 5ms
-    >>> delayed_spike = OutputDelayAt(neuron, 5.0 * u.ms)
-    >>> # Get the delayed value
-    >>> v_value = delayed_spike()
-    """
-
-    def __init__(
-        self,
-        module: Dynamics,
-        delay_time: Tuple,
-    ):
-        super().__init__()
-        assert isinstance(module, Dynamics), 'The module should be an instance of Dynamics.'
-        self.module = module
-        key = _get_output_delay_key()
-        if not module._has_after_update(key):
-            delay = Delay(jax.ShapeDtypeStruct(module.out_size, dtype=environ.dftype()), take_aware_unit=True)
-            module._add_after_update(key, receive_update_output(delay))
-        self.out_delay: Delay = module._get_after_update(key)
-        self.delay_info = self.out_delay.register_delay(*delay_time)
-
-    def __call__(self, *args, **kwargs):
-        return self.out_delay.retrieve_at_step(*self.delay_info)
-
-
-def _get_prefetch_delay_key(item) -> str:
-    return f'{item}-prefetch-delay'
-
-
-def _get_output_delay_key() -> str:
-    return f'output-delay'
-
-
-def _get_prefetch_item(target: Union[Prefetch, PrefetchDelayAt]) -> Any:
-    item = getattr(target.module, target.item, None)
-    if item is None:
-        raise AttributeError(f'The target {target.module} should have an `{target.item}` attribute.')
-    return item
-
-
-def _get_prefetch_item_delay(target: Union[Prefetch, PrefetchDelay, PrefetchDelayAt]) -> Delay:
-    assert isinstance(target.module, Dynamics), (
-        f'The target module should be an instance '
-        f'of Dynamics. But got {target.module}.'
-    )
-    delay = target.module._get_after_update(_get_prefetch_delay_key(target.item))
-    if not isinstance(delay, StateWithDelay):
-        raise TypeError(f'The prefetch target should be a {StateWithDelay.__name__} when accessing '
-                        f'its delay. But got {delay}.')
-    return delay
-
-
-def maybe_init_prefetch(target, *args, **kwargs):
-    """
-    Initialize a prefetch target if needed, based on its type.
-
-    This function ensures that prefetch references are properly initialized
-    and ready to use. It handles different types of prefetch objects by
-    performing the appropriate initialization action:
-    - For :py:class:`Prefetch` objects: retrieves the referenced item
-    - For :py:class:`PrefetchDelay` objects: retrieves the delay handler
-    - For :py:class:`PrefetchDelayAt` objects: registers the specified delay
-
-    Parameters
-    ----------
-    target : Union[Prefetch, PrefetchDelay, PrefetchDelayAt]
-        The prefetch target to initialize.
-    *args : Any
-        Additional positional arguments (unused).
-    **kwargs : Any
-        Additional keyword arguments (unused).
-
-    Returns
-    -------
-    None
-        This function performs initialization side effects only.
-
-    Notes
-    -----
-    This function is typically called internally when prefetched references
-    are used to ensure they are properly set up before access.
-    """
-    if isinstance(target, Prefetch):
-        _get_prefetch_item(target)
-
-    elif isinstance(target, PrefetchDelay):
-        _get_prefetch_item_delay(target)
-
-    elif isinstance(target, PrefetchDelayAt):
-        pass
-        # delay = _get_prefetch_item_delay(target)
-        # delay.register_delay(*target.delay_time)
-
-
-class DynamicsGroup(Module):
-    """
-    A group of :py:class:`~.Module` in which the updating order does not matter.
-
-    Args:
-      children_as_tuple: The children objects.
-      children_as_dict: The children objects.
-    """
-
-    __module__ = 'brainstate.nn'
-
-    if not TYPE_CHECKING:
-        def __init__(self, *children_as_tuple, **children_as_dict):
-            super().__init__()
-            self.layers_tuple = tuple(children_as_tuple)
-            self.layers_dict = dict(children_as_dict)
-
-    def update(self, *args, **kwargs):
-        """
-        Update function of a network.
-
-        In this update function, the update functions in children systems are iteratively called.
-        """
-        projs, dyns, others = self.nodes(allowed_hierarchy=(1, 1)).split(Projection, Dynamics)
-
-        # update nodes of projections
-        for node in projs.values():
-            node()
-
-        # update nodes of dynamics
-        for node in dyns.values():
-            node()
-
-        # update nodes with other types, including delays, ...
-        for node in others.values():
-            node()
-
-
-def receive_update_output(cls: object):
-    """
-    The decorator to mark the object (as the after updates) to receive the output of the update function.
-
-    That is, the `aft_update` will receive the return of the update function::
-
-      ret = model.update(*args, **kwargs)
-      for fun in model.aft_updates:
-        fun(ret)
-
-    """
-    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
-    if hasattr(cls, '_not_receive_update_output'):
-        delattr(cls, '_not_receive_update_output')
-    return cls
-
-
-def not_receive_update_output(cls: T) -> T:
-    """
-    The decorator to mark the object (as the after updates) to not receive the output of the update function.
-
-    That is, the `aft_update` will not receive the return of the update function::
-
-      ret = model.update(*args, **kwargs)
-      for fun in model.aft_updates:
-        fun()
-
-    """
-    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
-    cls._not_receive_update_output = True
-    return cls
-
-
-def receive_update_input(cls: object):
-    """
-    The decorator to mark the object (as the before updates) to receive the input of the update function.
-
-    That is, the `bef_update` will receive the input of the update function::
-
-
-      for fun in model.bef_updates:
-        fun(*args, **kwargs)
-      model.update(*args, **kwargs)
-
-    """
-    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
-    cls._receive_update_input = True
-    return cls
-
-
-def not_receive_update_input(cls: object):
-    """
-    The decorator to mark the object (as the before updates) to not receive the input of the update function.
-
-    That is, the `bef_update` will not receive the input of the update function::
-
-        for fun in model.bef_updates:
-          fun()
-        model.update()
-
-    """
-    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
-    if hasattr(cls, '_receive_update_input'):
-        delattr(cls, '_receive_update_input')
-    return cls
-
-
-def _input_label_start(label: str):
-    # unify the input label repr.
-    return f'{label} // '
-
-
-def _input_label_repr(name: str, label: Optional[str] = None):
-    # unify the input label repr.
-    return name if label is None else (_input_label_start(label) + str(name))
+# Copyright 2024 BDP Ecosystem Limited. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+
+# -*- coding: utf-8 -*-
+
+
+"""
+All the basic dynamics class for the ``brainstate``.
+
+For handling dynamical systems:
+
+- ``DynamicsGroup``: The class for a group of modules, which update ``Projection`` first,
+                   then ``Dynamics``, finally others.
+- ``Projection``: The class for the synaptic projection.
+- ``Dynamics``: The class for the dynamical system.
+
+For handling the delays:
+
+- ``Delay``: The class for all delays.
+- ``DelayAccess``: The class for the delay access.
+
+"""
+
+from typing import Any, Dict, Callable, Hashable, Optional, Union, TypeVar, TYPE_CHECKING, Tuple
+
+import jax
+import numpy as np
+
+from brainstate import environ
+from brainstate._state import State
+from brainstate.graph import Node
+from brainstate.mixin import ParamDescriber
+from brainstate.typing import Size, ArrayLike
+from ._delay import StateWithDelay, Delay
+from ._module import Module
+
+__all__ = [
+    'DynamicsGroup', 'Projection', 'Dynamics',
+    'Prefetch', 'PrefetchDelay', 'PrefetchDelayAt', 'OutputDelayAt',
+]
+
+T = TypeVar('T')
+_max_order = 10
+
+
+class Projection(Module):
+    """
+    Base class for synaptic projection modules in neural network modeling.
+
+    This class defines the interface for modules that handle projections between
+    neural populations. Projections process input signals and transform them
+    before they reach the target neurons, implementing the connectivity patterns
+    in neural networks.
+
+    In the BrainState execution order, Projection modules are updated before
+    Dynamics modules, following the natural information flow in neural systems:
+    1. Projections process inputs (synaptic transmission)
+    2. Dynamics update neuron states (neural integration)
+
+    The Projection class does not implement the update logic directly but delegates
+    to its child nodes. If no child nodes exist, it raises a ValueError.
+
+    Parameters
+    ----------
+    *args : Any
+        Arguments passed to the parent Module class.
+    **kwargs : Any
+        Keyword arguments passed to the parent Module class.
+
+    Raises
+    ------
+    ValueError
+        If the update() method is called but no child nodes are defined.
+
+    Notes
+    -----
+    Derived classes should implement specific projection behaviors, such as
+    dense connectivity, sparse connectivity, or specific weight update rules.
+    """
+    __module__ = 'brainstate.nn'
+
+    def update(self, *args, **kwargs):
+        sub_nodes = tuple(self.nodes(allowed_hierarchy=(1, 1)).values())
+        if len(sub_nodes):
+            for node in sub_nodes:
+                node(*args, **kwargs)
+        else:
+            raise ValueError('Do not implement the update() function.')
+
+
+class Dynamics(Module):
+    """
+    Base class for implementing neural dynamics models in BrainState.
+
+    Dynamics classes represent the core computational units in neural simulations,
+    implementing the differential equations or update rules that govern neural activity.
+    This class provides infrastructure for managing neural populations, handling inputs,
+    and coordinating updates within the simulation framework.
+
+    The Dynamics class serves several key purposes:
+    1. Managing neuron population geometry and size information
+    2. Handling current and delta (instantaneous change) inputs to neurons
+    3. Supporting before/after update hooks for computational dependencies
+    4. Providing access to delayed state variables through the prefetch mechanism
+    5. Establishing the execution order in neural network simulations
+
+    Parameters
+    ----------
+    in_size : Size
+        The geometry of the neuron population. Can be an integer (e.g., 10) for
+        1D neuron arrays, or a tuple (e.g., (10, 10)) for multi-dimensional populations.
+    name : Optional[str], default=None
+        Optional name identifier for this dynamics module.
+
+    Attributes
+    ----------
+    in_size : tuple
+        The shape/geometry of the neuron population.
+    out_size : tuple
+        The output shape, typically matches in_size.
+    current_inputs : Optional[Dict[str, Union[Callable, ArrayLike]]]
+        Dictionary of registered current input functions or arrays.
+    delta_inputs : Optional[Dict[str, Union[Callable, ArrayLike]]]
+        Dictionary of registered delta input functions or arrays.
+    before_updates : Optional[Dict[Hashable, Callable]]
+        Dictionary of functions to call before the main update.
+    after_updates : Optional[Dict[Hashable, Callable]]
+        Dictionary of functions to call after the main update.
+
+    Notes
+    -----
+    In the BrainState execution sequence, Dynamics modules are updated after
+    Projection modules and before other module types, reflecting the natural
+    flow of information in neural systems.
+
+    There are several essential attributes:
+
+    - ``size``: the geometry of the neuron group. For example, `(10, )` denotes a line of
+      neurons, `(10, 10)` denotes a neuron group aligned in a 2D space, `(10, 15, 4)` denotes
+      a 3-dimensional neuron group.
+    - ``num``: the flattened number of neurons in the group. For example, `size=(10, )` => \
+      `num=10`, `size=(10, 10)` => `num=100`, `size=(10, 15, 4)` => `num=600`.
+
+
+    See Also
+    --------
+    Module : Parent class providing base module functionality
+    Projection : Class for handling synaptic projections between neural populations
+    DynamicsGroup : Container for organizing multiple dynamics modules
+    """
+
+    __module__ = 'brainstate.nn'
+
+    graph_invisible_attrs = ('_before_updates', '_after_updates', '_current_inputs', '_delta_inputs')
+
+    # before updates
+    _before_updates: Optional[Dict[Hashable, Callable]]
+
+    # after updates
+    _after_updates: Optional[Dict[Hashable, Callable]]
+
+    # current inputs
+    _current_inputs: Optional[Dict[str, ArrayLike | Callable]]
+
+    # delta inputs
+    _delta_inputs: Optional[Dict[str, ArrayLike | Callable]]
+
+    def __init__(
+        self,
+        in_size: Size,
+        name: Optional[str] = None,
+    ):
+        # initialize
+        super().__init__(name=name)
+
+        # geometry size of neuron population
+        if isinstance(in_size, (list, tuple)):
+            if len(in_size) <= 0:
+                raise ValueError(f'"in_size" must be int, or a tuple/list of int. But we got {type(in_size)}')
+            if not isinstance(in_size[0], (int, np.integer)):
+                raise ValueError(f'"in_size" must be int, or a tuple/list of int. But we got {type(in_size)}')
+            in_size = tuple(in_size)
+        elif isinstance(in_size, (int, np.integer)):
+            in_size = (in_size,)
+        else:
+            raise ValueError(f'"in_size" must be int, or a tuple/list of int. But we got {type(in_size)}')
+        self.in_size = in_size
+
+        # current inputs
+        self._current_inputs = None
+
+        # delta inputs
+        self._delta_inputs = None
+
+        # before updates
+        self._before_updates = None
+
+        # after updates
+        self._after_updates = None
+
+        # in-/out- size of neuron population
+        self.out_size = self.in_size
+
+    # def __pretty_repr_item__(self, name, value):
+    #     if name in [
+    #         '_before_updates', '_after_updates', '_current_inputs', '_delta_inputs',
+    #         '_in_size', '_out_size', '_name', '_mode',
+    #     ]:
+    #         return (name, value) if value is None else (name[1:], value)  # skip the first `_`
+    #     return super().__pretty_repr_item__(name, value)
+
+    @property
+    def varshape(self):
+        """
+        Get the shape of variables in the neuron group.
+
+        This property provides access to the geometry (shape) of the neuron population,
+        which determines how variables and states are structured.
+
+        Returns
+        -------
+        tuple
+            A tuple representing the dimensional shape of the neuron group,
+            matching the in_size parameter provided during initialization.
+
+        See Also
+        --------
+        in_size : The input geometry specification for the neuron group
+        """
+        return self.in_size
+
+    @property
+    def current_inputs(self):
+        """
+        Get the dictionary of current inputs registered with this dynamics model.
+
+        Current inputs represent direct input currents that flow into the model.
+
+        Returns
+        -------
+        dict or None
+            A dictionary mapping keys to current input functions or values,
+            or None if no current inputs have been registered.
+
+        See Also
+        --------
+        add_current_input : Register a new current input
+        sum_current_inputs : Apply and sum all current inputs
+        delta_inputs : Dictionary of instantaneous change inputs
+        """
+        return self._current_inputs
+
+    @property
+    def delta_inputs(self):
+        """
+        Get the dictionary of delta inputs registered with this dynamics model.
+
+        Delta inputs represent instantaneous changes to state variables (dX/dt).
+
+        Returns
+        -------
+        dict or None
+            A dictionary mapping keys to delta input functions or values,
+            or None if no delta inputs have been registered.
+
+        See Also
+        --------
+        add_delta_input : Register a new delta input
+        sum_delta_inputs : Apply and sum all delta inputs
+        current_inputs : Dictionary of direct current inputs
+        """
+        return self._delta_inputs
+
+    def add_current_input(
+        self,
+        key: str,
+        inp: Union[Callable, ArrayLike],
+        label: Optional[str] = None
+    ):
+        """
+        Add a current input function or array to the dynamics model.
+
+        Current inputs represent direct input currents that can be accessed during
+        model updates through the `sum_current_inputs()` method.
+
+        Parameters
+        ----------
+        key : str
+            Unique identifier for this current input. Used to retrieve or reference
+            the input later.
+        inp : Union[Callable, ArrayLike]
+            The input data or function that generates input data.
+            - If callable: Will be called during updates with arguments passed to `sum_current_inputs()`
+            - If array-like: Will be applied once and then automatically removed from available inputs
+        label : Optional[str], default=None
+            Optional grouping label for the input. When provided, allows selective
+            processing of inputs by label in `sum_current_inputs()`.
+
+        Raises
+        ------
+        ValueError
+            If the key has already been used for a different current input.
+
+        Notes
+        -----
+        - Inputs with the same label can be processed together using the `label`
+          parameter in `sum_current_inputs()`.
+        - Non-callable inputs are consumed when used (removed after first use).
+        - Callable inputs persist and can be called repeatedly.
+
+        See Also
+        --------
+        sum_current_inputs : Sum all current inputs matching a given label
+        add_delta_input : Add a delta input function or array
+        """
+        key = _input_label_repr(key, label)
+        if self._current_inputs is None:
+            self._current_inputs = dict()
+        if key in self._current_inputs:
+            if id(self._current_inputs[key]) != id(inp):
+                raise ValueError(f'Key "{key}" has been defined and used in the current inputs of {self}.')
+        self._current_inputs[key] = inp
+
+    def add_delta_input(
+        self,
+        key: str,
+        inp: Union[Callable, ArrayLike],
+        label: Optional[str] = None
+    ):
+        """
+        Add a delta input function or array to the dynamics model.
+
+        Delta inputs represent instantaneous changes to the model state (i.e., dX/dt contributions).
+        This method registers a function or array that provides delta inputs which will be
+        accessible during model updates through the `sum_delta_inputs()` method.
+
+        Parameters
+        ----------
+        key : str
+            Unique identifier for this delta input. Used to retrieve or reference
+            the input later.
+        inp : Union[Callable, ArrayLike]
+            The input data or function that generates input data.
+            - If callable: Will be called during updates with arguments passed to `sum_delta_inputs()`
+            - If array-like: Will be applied once and then automatically removed from available inputs
+        label : Optional[str], default=None
+            Optional grouping label for the input. When provided, allows selective
+            processing of inputs by label in `sum_delta_inputs()`.
+
+        Raises
+        ------
+        ValueError
+            If the key has already been used for a different delta input.
+
+        Notes
+        -----
+        - Inputs with the same label can be processed together using the `label`
+          parameter in `sum_delta_inputs()`.
+        - Non-callable inputs are consumed when used (removed after first use).
+        - Callable inputs persist and can be called repeatedly.
+
+        See Also
+        --------
+        sum_delta_inputs : Sum all delta inputs matching a given label
+        add_current_input : Add a current input function or array
+        """
+        key = _input_label_repr(key, label)
+        if self._delta_inputs is None:
+            self._delta_inputs = dict()
+        if key in self._delta_inputs:
+            if id(self._delta_inputs[key]) != id(inp):
+                raise ValueError(f'Key "{key}" has been defined and used.')
+        self._delta_inputs[key] = inp
+
+    def get_input(self, key: str):
+        """
+        Get a registered input function by its key.
+
+        Retrieves either a current input or a delta input function that was previously
+        registered with the given key. This method checks both current_inputs and
+        delta_inputs dictionaries for the specified key.
+
+        Parameters
+        ----------
+        key : str
+            The unique identifier used when the input function was registered.
+
+        Returns
+        -------
+        Callable or ArrayLike
+            The input function or array associated with the given key.
+
+        Raises
+        ------
+        ValueError
+            If no input function is found with the specified key in either
+            current_inputs or delta_inputs.
+
+        See Also
+        --------
+        add_current_input : Register a current input function
+        add_delta_input : Register a delta input function
+
+        Examples
+        --------
+        >>> model = Dynamics(10)
+        >>> model.add_current_input('stimulus', lambda t: np.sin(t))
+        >>> input_func = model.get_input('stimulus')
+        >>> input_func(0.5)  # Returns sin(0.5)
+        """
+        if self._current_inputs is not None and key in self._current_inputs:
+            return self._current_inputs[key]
+        elif self._delta_inputs is not None and key in self._delta_inputs:
+            return self._delta_inputs[key]
+        else:
+            raise ValueError(f'Input key {key} is not in current/delta inputs of the module {self}.')
+
+    def sum_current_inputs(
+        self,
+        init: Any,
+        *args,
+        label: Optional[str] = None,
+        **kwargs
+    ):
+        """
+        Summarize all current inputs by applying and summing all registered current input functions.
+
+        This method iterates through all registered current input functions (from `.current_inputs`)
+        and applies them to calculate the total input current for the dynamics model. It adds all results
+        to the initial value provided.
+
+        Parameters
+        ----------
+        init : Any
+            The initial value to which all current inputs will be added.
+        *args : tuple
+            Variable length argument list passed to each current input function.
+        label : Optional[str], default=None
+            If provided, only process current inputs with this label prefix.
+            When None, process all current inputs regardless of label.
+        **kwargs : dict
+            Arbitrary keyword arguments passed to each current input function.
+
+        Returns
+        -------
+        Any
+            The initial value plus all applicable current inputs summed together.
+
+        Notes
+        -----
+        - Non-callable current inputs are applied once and then automatically removed from
+          the current_inputs dictionary.
+        - Callable current inputs remain registered for subsequent calls.
+        - When a label is provided, only current inputs with keys starting with that label
+          are applied.
+        """
+        if self._current_inputs is None:
+            return init
+        if label is None:
+            filter_fn = lambda k: True
+        else:
+            label_repr = _input_label_start(label)
+            filter_fn = lambda k: k.startswith(label_repr)
+        for key in tuple(self._current_inputs.keys()):
+            if filter_fn(key):
+                out = self._current_inputs[key]
+                if callable(out):
+                    try:
+                        init = init + out(*args, **kwargs)
+                    except Exception as e:
+                        raise ValueError(
+                            f'Error in delta input value {key}: {out}\n'
+                            f'Error: {e}'
+                        ) from e
+                else:
+                    try:
+                        init = init + out
+                    except Exception as e:
+                        raise ValueError(
+                            f'Error in delta input value {key}: {out}\n'
+                            f'Error: {e}'
+                        ) from e
+                    self._current_inputs.pop(key)
+        return init
+
+    def sum_delta_inputs(
+        self,
+        init: Any,
+        *args,
+        label: Optional[str] = None,
+        **kwargs
+    ):
+        """
+        Summarize all delta inputs by applying and summing all registered delta input functions.
+
+        This method iterates through all registered delta input functions (from `.delta_inputs`)
+        and applies them to calculate instantaneous changes to model states. It adds all results
+        to the initial value provided.
+
+        Parameters
+        ----------
+        init : Any
+            The initial value to which all delta inputs will be added.
+        *args : tuple
+            Variable length argument list passed to each delta input function.
+        label : Optional[str], default=None
+            If provided, only process delta inputs with this label prefix.
+            When None, process all delta inputs regardless of label.
+        **kwargs : dict
+            Arbitrary keyword arguments passed to each delta input function.
+
+        Returns
+        -------
+        Any
+            The initial value plus all applicable delta inputs summed together.
+
+        Notes
+        -----
+        - Non-callable delta inputs are applied once and then automatically removed from
+          the delta_inputs dictionary.
+        - Callable delta inputs remain registered for subsequent calls.
+        - When a label is provided, only delta inputs with keys starting with that label
+          are applied.
+        """
+        if self._delta_inputs is None:
+            return init
+        if label is None:
+            filter_fn = lambda k: True
+        else:
+            label_repr = _input_label_start(label)
+            filter_fn = lambda k: k.startswith(label_repr)
+        for key in tuple(self._delta_inputs.keys()):
+            if filter_fn(key):
+                out = self._delta_inputs[key]
+                if callable(out):
+                    try:
+                        init = init + out(*args, **kwargs)
+                    except Exception as e:
+                        raise ValueError(
+                            f'Error in delta input function {key}: {out}\n'
+                            f'Error: {e}'
+                        ) from e
+                else:
+                    try:
+                        init = init + out
+                    except Exception as e:
+                        raise ValueError(
+                            f'Error in delta input value {key}: {out}\n'
+                            f'Error: {e}'
+                        ) from e
+                    self._delta_inputs.pop(key)
+        return init
+
+    @property
+    def before_updates(self):
+        """
+        Get the dictionary of functions to execute before the module's update.
+
+        Returns
+        -------
+        dict or None
+            Dictionary mapping keys to callable functions that will be executed
+            before the main update, or None if no before updates are registered.
+
+        Notes
+        -----
+        Before updates are executed in the order they were registered whenever
+        the module is called via __call__.
+        """
+        return self._before_updates
+
+    @property
+    def after_updates(self):
+        """
+        Get the dictionary of functions to execute after the module's update.
+
+        Returns
+        -------
+        dict or None
+            Dictionary mapping keys to callable functions that will be executed
+            after the main update, or None if no after updates are registered.
+
+        Notes
+        -----
+        After updates are executed in the order they were registered whenever
+        the module is called via __call__, and may optionally receive the return
+        value from the update method.
+        """
+        return self._after_updates
+
+    def _add_before_update(self, key: Any, fun: Callable):
+        """
+        Register a function to be executed before the module's update.
+
+        Parameters
+        ----------
+        key : Any
+            A unique identifier for the update function.
+        fun : Callable
+            The function to execute before the module's update.
+
+        Raises
+        ------
+        KeyError
+            If the key is already registered in before_updates.
+
+        Notes
+        -----
+        Internal method used by the module system to register dependencies.
+        """
+        if self._before_updates is None:
+            self._before_updates = dict()
+        if key in self.before_updates:
+            raise KeyError(f'{key} has been registered in before_updates of {self}')
+        self.before_updates[key] = fun
+
+    def _add_after_update(self, key: Any, fun: Callable):
+        """
+        Register a function to be executed after the module's update.
+
+        Parameters
+        ----------
+        key : Any
+            A unique identifier for the update function.
+        fun : Callable
+            The function to execute after the module's update.
+
+        Raises
+        ------
+        KeyError
+            If the key is already registered in after_updates.
+
+        Notes
+        -----
+        Internal method used by the module system to register dependencies.
+        """
+        if self._after_updates is None:
+            self._after_updates = dict()
+        if key in self.after_updates:
+            raise KeyError(f'{key} has been registered in after_updates of {self}')
+        self.after_updates[key] = fun
+
+    def _get_before_update(self, key: Any):
+        """
+        Retrieve a registered before-update function by its key.
+
+        Parameters
+        ----------
+        key : Any
+            The identifier of the before-update function to retrieve.
+
+        Returns
+        -------
+        Callable
+            The registered before-update function.
+
+        Raises
+        ------
+        KeyError
+            If the key is not registered in before_updates or if before_updates is None.
+        """
+        if self._before_updates is None:
+            raise KeyError(f'{key} is not registered in before_updates of {self}')
+        if key not in self.before_updates:
+            raise KeyError(f'{key} is not registered in before_updates of {self}')
+        return self.before_updates.get(key)
+
+    def _get_after_update(self, key: Any):
+        """
+        Retrieve a registered after-update function by its key.
+
+        Parameters
+        ----------
+        key : Any
+            The identifier of the after-update function to retrieve.
+
+        Returns
+        -------
+        Callable
+            The registered after-update function.
+
+        Raises
+        ------
+        KeyError
+            If the key is not registered in after_updates or if after_updates is None.
+        """
+        if self._after_updates is None:
+            raise KeyError(f'{key} is not registered in after_updates of {self}')
+        if key not in self.after_updates:
+            raise KeyError(f'{key} is not registered in after_updates of {self}')
+        return self.after_updates.get(key)
+
+    def _has_before_update(self, key: Any):
+        """
+        Check if a before-update function is registered with the given key.
+
+        Parameters
+        ----------
+        key : Any
+            The identifier to check for in the before_updates dictionary.
+
+        Returns
+        -------
+        bool
+            True if the key is registered in before_updates, False otherwise.
+        """
+        if self._before_updates is None:
+            return False
+        return key in self.before_updates
+
+    def _has_after_update(self, key: Any):
+        """
+        Check if an after-update function is registered with the given key.
+
+        Parameters
+        ----------
+        key : Any
+            The identifier to check for in the after_updates dictionary.
+
+        Returns
+        -------
+        bool
+            True if the key is registered in after_updates, False otherwise.
+        """
+        if self._after_updates is None:
+            return False
+        return key in self.after_updates
+
+    def __call__(self, *args, **kwargs):
+        """
+        The shortcut to call ``update`` methods.
+        """
+
+        # ``before_updates``
+        if self.before_updates is not None:
+            for model in self.before_updates.values():
+                if hasattr(model, '_receive_update_input'):
+                    model(*args, **kwargs)
+                else:
+                    model()
+
+        # update the model self
+        ret = self.update(*args, **kwargs)
+
+        # ``after_updates``
+        if self.after_updates is not None:
+            for model in self.after_updates.values():
+                if hasattr(model, '_not_receive_update_output'):
+                    model()
+                else:
+                    model(ret)
+        return ret
+
+    def prefetch(self, item: str) -> 'Prefetch':
+        """
+        Create a reference to a state or variable that may not be initialized yet.
+
+        This method allows accessing module attributes or states before they are
+        fully defined, acting as a placeholder that will be resolved when called.
+        Particularly useful for creating references to variables that will be defined
+        during initialization or runtime.
+
+        Parameters
+        ----------
+        item : str
+            The name of the attribute or state to reference.
+
+        Returns
+        -------
+        Prefetch
+            A Prefetch object that provides access to the referenced item.
+
+        Examples
+        --------
+        >>> import brainstate
+        >>> import brainunit as u
+        >>> neuron = brainstate.nn.LIF(...)
+        >>> v_ref = neuron.prefetch('V')  # Reference to voltage
+        >>> v_value = v_ref()  # Get current value
+        >>> delayed_v = v_ref.delay.at(5.0 * u.ms)  # Get delayed value
+        """
+        return Prefetch(self, item)
+
+    def align_pre(self, dyn: Union[ParamDescriber[T], T]) -> T:
+        """
+        Registers a dynamics module to execute after this module.
+
+        This method establishes a sequential execution relationship where the specified
+        dynamics module will be called after this module completes its update. This
+        creates a feed-forward connection in the computational graph.
+
+        Parameters
+        ----------
+        dyn : Union[ParamDescriber[T], T]
+            The dynamics module to be executed after this module. Can be either:
+            - An instance of Dynamics
+            - A ParamDescriber that can instantiate a Dynamics object
+
+        Returns
+        -------
+        T
+            The dynamics module that was registered, allowing for method chaining.
+
+        Raises
+        ------
+        TypeError
+            If the input is not a Dynamics instance or a ParamDescriber that creates
+            a Dynamics instance.
+
+        Examples
+        --------
+        >>> import brainstate
+        >>> n1 = brainstate.nn.LIF(10)
+        >>> n1.align_pre(brainstate.nn.Expon.desc(n1.varshape))  # n2 will run after n1
+        """
+        if isinstance(dyn, Dynamics):
+            self._add_after_update(id(dyn), dyn)
+            return dyn
+        elif isinstance(dyn, ParamDescriber):
+            if not issubclass(dyn.cls, Dynamics):
+                raise TypeError(f'The input {dyn} should be an instance of {Dynamics}.')
+            if not self._has_after_update(dyn.identifier):
+                self._add_after_update(
+                    dyn.identifier,
+                    dyn() if ('in_size' in dyn.kwargs or len(dyn.args) > 0) else dyn(in_size=self.varshape)
+                )
+            return self._get_after_update(dyn.identifier)
+        else:
+            raise TypeError(f'The input {dyn} should be an instance of {Dynamics} or a delayed initializer.')
+
+    def prefetch_delay(self, state: str, delay_time, init: Callable = None) -> 'PrefetchDelayAt':
+        """
+        Create a reference to a delayed state or variable in the module.
+
+        This method simplifies the process of accessing a delayed version of a state or variable
+        within the module. It first creates a prefetch reference to the specified state,
+        then specifies the delay time for accessing this state.
+
+        Args:
+            state (str): The name of the state or variable to reference.
+            delay_time (ArrayLike): The amount of time to delay the variable access,
+                typically in time units (e.g., milliseconds).
+            init (Callable, optional): An optional initialization function to provide
+                a default value if the delayed state is not yet available.
+
+        Returns:
+            PrefetchDelayAt: An object that provides access to the variable at the specified delay time.
+        """
+        return PrefetchDelayAt(self, state, delay_time, init=init)
+
+    def output_delay(self, *delay_time) -> 'OutputDelayAt':
+        """
+        Create a reference to the delayed output of the module.
+
+        This method simplifies the process of accessing a delayed version of the module's output.
+        It instantiates an `OutputDelayAt` object, which can be used to retrieve the output value
+        at the specified delay time.
+
+        Args:
+            delay (Optional[ArrayLike]): The amount of time to delay the output access,
+                typically in time units (e.g., milliseconds). Defaults to None.
+
+        Returns:
+            OutputDelayAt: An object that provides access to the module's output at the specified delay time.
+        """
+        return OutputDelayAt(self, delay_time)
+
+
+class Prefetch(Node):
+    """
+    Prefetch a state or variable in a module before it is initialized.
+
+
+    This class provides a mechanism to reference a module's state or attribute
+    that may not have been initialized yet. It acts as a placeholder or reference
+    that will be resolved when called.
+
+    Use cases:
+    - Access variables within dynamics modules that will be defined later
+    - Create references to states across module boundaries
+    - Enable access to delayed states through the `.delay` property
+
+    Parameters
+    ----------
+    module : Module
+        The module that contains or will contain the referenced item.
+    item : str
+        The attribute name of the state or variable to prefetch.
+
+    Examples
+    --------
+    >>> import brainstate
+    >>> import brainunit as u
+    >>> neuron = brainstate.nn.LIF(...)
+    >>> v_reference = neuron.prefetch('V')  # Reference to voltage before initialization
+    >>> v_value = v_reference()  # Get the current value
+    >>> delay_ref = v_reference.delay.at(5.0 * u.ms)  # Reference voltage delayed by 5ms
+
+    Notes
+    -----
+    When called, this class retrieves the current value of the referenced item.
+    Use the `.delay` property to access delayed versions of the state.
+
+    """
+
+    def __init__(self, module: Dynamics, item: str):
+        """
+        Initialize a Prefetch object.
+
+        Parameters
+        ----------
+        module : Module
+            The module that contains or will contain the referenced item.
+        item : str
+            The attribute name of the state or variable to prefetch.
+        """
+        super().__init__()
+        self.module = module
+        self.item = item
+
+    @property
+    def delay(self):
+        """
+        Access delayed versions of the prefetched item.
+
+        Returns
+        -------
+        PrefetchDelay
+            An object that provides access to delayed versions of the prefetched item.
+        """
+        return PrefetchDelay(self.module, self.item)
+        # return PrefetchDelayAt(self.module, self.item, time)
+
+    def __call__(self, *args, **kwargs):
+        """
+        Get the current value of the prefetched item.
+
+        Returns
+        -------
+        Any
+            The current value of the referenced item. If the item is a State object,
+            returns its value attribute, otherwise returns the item itself.
+        """
+        item = _get_prefetch_item(self)
+        return item.value if isinstance(item, State) else item
+
+    def get_item_value(self):
+        """
+        Get the current value of the prefetched item.
+
+        Similar to __call__, but explicitly named for clarity.
+
+        Returns
+        -------
+        Any
+            The current value of the referenced item. If the item is a State object,
+            returns its value attribute, otherwise returns the item itself.
+        """
+        item = _get_prefetch_item(self)
+        return item.value if isinstance(item, State) else item
+
+    def get_item(self):
+        """
+        Get the referenced item object itself, not its value.
+
+        Returns
+        -------
+        Any
+            The actual referenced item from the module, which could be a State
+            object or any other attribute.
+        """
+        return _get_prefetch_item(self)
+
+
+class PrefetchDelay(Node):
+    """
+    Provides access to delayed versions of a prefetched state or variable.
+
+    This class acts as an intermediary for accessing delayed values of module variables.
+    It doesn't retrieve values directly but provides methods to specify the delay time
+    via the `at()` method.
+
+    Parameters
+    ----------
+    module : Dynamics
+        The dynamics module that contains the referenced state or variable.
+    item : str
+        The name of the state or variable to access with delay.
+
+    Examples
+    --------
+    >>> import brainstate
+    >>> import brainunit as u
+    >>> neuron = brainstate.nn.LIF(10)
+    >>> # Access voltage delayed by 5ms
+    >>> delayed_v = neuron.prefetch('V').delay.at(5.0 * u.ms)
+    >>> delayed_value = delayed_v()  # Get the delayed value
+    """
+
+    def __init__(self, module: Dynamics, item: str):
+        self.module = module
+        self.item = item
+
+    def at(self, *delay_time):
+        """
+        Specifies the delay time for accessing the variable.
+
+        Parameters
+        ----------
+        time : ArrayLike
+            The amount of time to delay the variable access, typically in time units
+            (e.g., milliseconds).
+
+        Returns
+        -------
+        PrefetchDelayAt
+            An object that provides access to the variable at the specified delay time.
+        """
+        return PrefetchDelayAt(self.module, self.item, delay_time)
+
+
+class PrefetchDelayAt(Node):
+    """
+    Provides access to a specific delayed state or variable value at the specific time.
+
+    This class represents the final step in the prefetch delay chain, providing
+    actual access to state values at a specific delay time. It converts the
+    specified time delay into steps and registers the delay with the appropriate
+    StateWithDelay handler.
+
+    Parameters
+    ----------
+    module : Dynamics
+        The dynamics module that contains the referenced state or variable.
+    item : str
+        The name of the state or variable to access with delay.
+    time : ArrayLike
+        The amount of time to delay access by, typically in time units (e.g., milliseconds).
+
+    Examples
+    --------
+    >>> import brainstate
+    >>> import brainunit as u
+    >>> neuron = brainstate.nn.LIF(10)
+    >>> # Create a reference to voltage delayed by 5ms
+    >>> delayed_v = PrefetchDelayAt(neuron, 'V', 5.0 * u.ms)
+    >>> # Get the delayed value
+    >>> v_value = delayed_v()
+    """
+
+    def __init__(
+        self,
+        module: Dynamics,
+        item: str,
+        delay_time: Tuple,
+        init: Callable = None
+    ):
+        """
+        Initialize a PrefetchDelayAt object.
+
+        Parameters
+        ----------
+        module : Dynamics
+            The dynamics module that contains the referenced state or variable.
+        item : str
+            The name of the state or variable to access with delay.
+        delay_time : Tuple
+            The amount of time to delay access by, typically in time units (e.g., milliseconds).
+        """
+        super().__init__()
+        assert isinstance(module, Dynamics), 'The module should be an instance of Dynamics.'
+        self.module = module
+        self.item = item
+        if not isinstance(delay_time, (tuple, list)):
+            delay_time = (delay_time,)
+        self.delay_time = delay_time
+        if len(delay_time) > 0:
+            key = _get_prefetch_delay_key(item)
+            if not module._has_after_update(key):
+                module._add_after_update(
+                    key,
+                    not_receive_update_output(
+                        StateWithDelay(module, item, init=init)
+                    )
+                )
+            self.state_delay: StateWithDelay = module._get_after_update(key)
+            self.delay_info = self.state_delay.register_delay(*delay_time)
+
+    def __call__(self, *args, **kwargs):
+        """
+        Retrieve the value of the state at the specified delay time.
+
+        Returns
+        -------
+        Any
+            The value of the state or variable at the specified delay time.
+        """
+        if len(self.delay_time) == 0:
+            return _get_prefetch_item(self).value
+        else:
+            return self.state_delay.retrieve_at_step(*self.delay_info)
+
+
+class OutputDelayAt(Node):
+    """
+    Provides access to a specific delayed state or variable value at the specific time.
+
+    This class represents the final step in the prefetch delay chain, providing
+    actual access to state values at a specific delay time. It converts the
+    specified time delay into steps and registers the delay with the appropriate
+    StateWithDelay handler.
+
+    Parameters
+    ----------
+    module : Dynamics
+        The dynamics module that contains the referenced state or variable.
+    time : ArrayLike
+        The amount of time to delay access by, typically in time units (e.g., milliseconds).
+
+    Examples
+    --------
+    >>> import brainstate
+    >>> import brainunit as u
+    >>> neuron = brainstate.nn.LIF(10)
+    >>> # Create a reference to voltage delayed by 5ms
+    >>> delayed_spike = OutputDelayAt(neuron, 5.0 * u.ms)
+    >>> # Get the delayed value
+    >>> v_value = delayed_spike()
+    """
+
+    def __init__(
+        self,
+        module: Dynamics,
+        delay_time: Tuple,
+    ):
+        super().__init__()
+        assert isinstance(module, Dynamics), 'The module should be an instance of Dynamics.'
+        self.module = module
+        key = _get_output_delay_key()
+        if not module._has_after_update(key):
+            delay = Delay(jax.ShapeDtypeStruct(module.out_size, dtype=environ.dftype()), take_aware_unit=True)
+            module._add_after_update(key, receive_update_output(delay))
+        self.out_delay: Delay = module._get_after_update(key)
+        self.delay_info = self.out_delay.register_delay(*delay_time)
+
+    def __call__(self, *args, **kwargs):
+        return self.out_delay.retrieve_at_step(*self.delay_info)
+
+
+def _get_prefetch_delay_key(item) -> str:
+    return f'{item}-prefetch-delay'
+
+
+def _get_output_delay_key() -> str:
+    return f'output-delay'
+
+
+def _get_prefetch_item(target: Union[Prefetch, PrefetchDelayAt]) -> Any:
+    item = getattr(target.module, target.item, None)
+    if item is None:
+        raise AttributeError(f'The target {target.module} should have an `{target.item}` attribute.')
+    return item
+
+
+def _get_prefetch_item_delay(target: Union[Prefetch, PrefetchDelay, PrefetchDelayAt]) -> Delay:
+    assert isinstance(target.module, Dynamics), (
+        f'The target module should be an instance '
+        f'of Dynamics. But got {target.module}.'
+    )
+    delay = target.module._get_after_update(_get_prefetch_delay_key(target.item))
+    if not isinstance(delay, StateWithDelay):
+        raise TypeError(f'The prefetch target should be a {StateWithDelay.__name__} when accessing '
+                        f'its delay. But got {delay}.')
+    return delay
+
+
+def maybe_init_prefetch(target, *args, **kwargs):
+    """
+    Initialize a prefetch target if needed, based on its type.
+
+    This function ensures that prefetch references are properly initialized
+    and ready to use. It handles different types of prefetch objects by
+    performing the appropriate initialization action:
+    - For :py:class:`Prefetch` objects: retrieves the referenced item
+    - For :py:class:`PrefetchDelay` objects: retrieves the delay handler
+    - For :py:class:`PrefetchDelayAt` objects: registers the specified delay
+
+    Parameters
+    ----------
+    target : Union[Prefetch, PrefetchDelay, PrefetchDelayAt]
+        The prefetch target to initialize.
+    *args : Any
+        Additional positional arguments (unused).
+    **kwargs : Any
+        Additional keyword arguments (unused).
+
+    Returns
+    -------
+    None
+        This function performs initialization side effects only.
+
+    Notes
+    -----
+    This function is typically called internally when prefetched references
+    are used to ensure they are properly set up before access.
+    """
+    if isinstance(target, Prefetch):
+        _get_prefetch_item(target)
+
+    elif isinstance(target, PrefetchDelay):
+        _get_prefetch_item_delay(target)
+
+    elif isinstance(target, PrefetchDelayAt):
+        pass
+        # delay = _get_prefetch_item_delay(target)
+        # delay.register_delay(*target.delay_time)
+
+
+class DynamicsGroup(Module):
+    """
+    A group of :py:class:`~.Module` in which the updating order does not matter.
+
+    Args:
+      children_as_tuple: The children objects.
+      children_as_dict: The children objects.
+    """
+
+    __module__ = 'brainstate.nn'
+
+    if not TYPE_CHECKING:
+        def __init__(self, *children_as_tuple, **children_as_dict):
+            super().__init__()
+            self.layers_tuple = tuple(children_as_tuple)
+            self.layers_dict = dict(children_as_dict)
+
+    def update(self, *args, **kwargs):
+        """
+        Update function of a network.
+
+        In this update function, the update functions in children systems are iteratively called.
+        """
+        projs, dyns, others = self.nodes(allowed_hierarchy=(1, 1)).split(Projection, Dynamics)
+
+        # update nodes of projections
+        for node in projs.values():
+            node()
+
+        # update nodes of dynamics
+        for node in dyns.values():
+            node()
+
+        # update nodes with other types, including delays, ...
+        for node in others.values():
+            node()
+
+
+def receive_update_output(cls: object):
+    """
+    The decorator to mark the object (as the after updates) to receive the output of the update function.
+
+    That is, the `aft_update` will receive the return of the update function::
+
+      ret = model.update(*args, **kwargs)
+      for fun in model.aft_updates:
+        fun(ret)
+
+    """
+    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
+    if hasattr(cls, '_not_receive_update_output'):
+        delattr(cls, '_not_receive_update_output')
+    return cls
+
+
+def not_receive_update_output(cls: T) -> T:
+    """
+    The decorator to mark the object (as the after updates) to not receive the output of the update function.
+
+    That is, the `aft_update` will not receive the return of the update function::
+
+      ret = model.update(*args, **kwargs)
+      for fun in model.aft_updates:
+        fun()
+
+    """
+    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
+    cls._not_receive_update_output = True
+    return cls
+
+
+def receive_update_input(cls: object):
+    """
+    The decorator to mark the object (as the before updates) to receive the input of the update function.
+
+    That is, the `bef_update` will receive the input of the update function::
+
+
+      for fun in model.bef_updates:
+        fun(*args, **kwargs)
+      model.update(*args, **kwargs)
+
+    """
+    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
+    cls._receive_update_input = True
+    return cls
+
+
+def not_receive_update_input(cls: object):
+    """
+    The decorator to mark the object (as the before updates) to not receive the input of the update function.
+
+    That is, the `bef_update` will not receive the input of the update function::
+
+        for fun in model.bef_updates:
+          fun()
+        model.update()
+
+    """
+    # assert isinstance(cls, Module), 'The input class should be instance of Module.'
+    if hasattr(cls, '_receive_update_input'):
+        delattr(cls, '_receive_update_input')
+    return cls
+
+
+def _input_label_start(label: str):
+    # unify the input label repr.
+    return f'{label} // '
+
+
+def _input_label_repr(name: str, label: Optional[str] = None):
+    # unify the input label repr.
+    return name if label is None else (_input_label_start(label) + str(name))