diff --git a/captum/_utils/av.py b/captum/_utils/av.py
index 2ab4ae268..97329a8b1 100644
--- a/captum/_utils/av.py
+++ b/captum/_utils/av.py
@@ -351,8 +351,7 @@ def _compute_and_save_activations(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         identifier: str,
         num_id: str,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         load_from_disk: bool = True,
     ) -> None:
         r"""
diff --git a/captum/_utils/common.py b/captum/_utils/common.py
index 12784bfaf..6459cd8aa 100644
--- a/captum/_utils/common.py
+++ b/captum/_utils/common.py
@@ -273,9 +273,25 @@ def _format_float_or_tensor_into_tuples(
     return inputs
 
 
+@overload
+def _format_additional_forward_args(
+    # pyre-fixme[24]: Generic type `tuple` expects at least 1 type parameter.
+    additional_forward_args: Union[Tensor, Tuple]
+    # pyre-fixme[24]: Generic type `tuple` expects at least 1 type parameter.
+) -> Tuple: ...
+
+
+@overload
+def _format_additional_forward_args(  # type: ignore
+    additional_forward_args: Optional[object],
+    # pyre-fixme[24]: Generic type `tuple` expects at least 1 type parameter.
+) -> Union[None, Tuple]: ...
+
+
 def _format_additional_forward_args(
     additional_forward_args: Optional[object],
-) -> Union[None, Tuple[object, ...]]:
+    # pyre-fixme[24]: Generic type `tuple` expects at least 1 type parameter.
+) -> Union[None, Tuple]:
     if additional_forward_args is not None and not isinstance(
         additional_forward_args, tuple
     ):
@@ -284,8 +300,8 @@ def _format_additional_forward_args(
 
 
 def _expand_additional_forward_args(
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any,
+    # pyre-fixme[24]: Generic type `tuple` expects at least 1 type parameter.
+    additional_forward_args: Union[None, Tuple],
     n_steps: int,
     expansion_type: ExpansionTypes = ExpansionTypes.repeat,
     # pyre-fixme[24]: Generic type `tuple` expects at least 1 type parameter.
@@ -557,8 +573,7 @@ def _run_forward(
     # pyre-fixme[2]: Parameter annotation cannot be `Any`.
     inputs: Any,
     target: TargetType = None,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
 ) -> Union[Tensor, Future[Tensor]]:
     forward_func_args = signature(forward_func).parameters
     if len(forward_func_args) == 0:
diff --git a/captum/_utils/gradient.py b/captum/_utils/gradient.py
index 2dab8154d..69502b744 100644
--- a/captum/_utils/gradient.py
+++ b/captum/_utils/gradient.py
@@ -104,8 +104,7 @@ def compute_gradients(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     target_ind: TargetType = None,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
 ) -> Tuple[Tensor, ...]:
     r"""
     Computes gradients of the output with respect to inputs for an
@@ -175,8 +174,7 @@ def _forward_layer_eval(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: List[Module],
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     device_ids: Union[None, List[int]] = None,
     attribute_to_layer_input: bool = False,
     grad_enabled: bool = False,
@@ -191,8 +189,7 @@ def _forward_layer_eval(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: Module,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     device_ids: Union[None, List[int]] = None,
     attribute_to_layer_input: bool = False,
     grad_enabled: bool = False,
@@ -204,7 +201,7 @@ def _forward_layer_eval(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: ModuleOrModuleList,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     device_ids: Union[None, List[int]] = None,
     attribute_to_layer_input: bool = False,
     grad_enabled: bool = False,
@@ -233,8 +230,7 @@ def _forward_layer_distributed_eval(
     inputs: Any,
     layer: ModuleOrModuleList,
     target_ind: TargetType = None,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     attribute_to_layer_input: bool = False,
     forward_hook_with_return: Literal[False] = False,
     require_layer_grads: bool = False,
@@ -250,7 +246,7 @@ def _forward_layer_distributed_eval(
     inputs: Any,
     layer: ModuleOrModuleList,
     target_ind: TargetType = None,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     attribute_to_layer_input: bool = False,
     *,
     forward_hook_with_return: Literal[True],
@@ -264,7 +260,7 @@ def _forward_layer_distributed_eval(
     inputs: Any,
     layer: ModuleOrModuleList,
     target_ind: TargetType = None,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     attribute_to_layer_input: bool = False,
     forward_hook_with_return: bool = False,
     require_layer_grads: bool = False,
@@ -427,8 +423,7 @@ def _forward_layer_eval_with_neuron_grads(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: Module,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     *,
     # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
     gradient_neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
@@ -446,7 +441,7 @@ def _forward_layer_eval_with_neuron_grads(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: List[Module],
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     gradient_neuron_selector: None = None,
     grad_enabled: bool = False,
     device_ids: Union[None, List[int]] = None,
@@ -462,7 +457,7 @@ def _forward_layer_eval_with_neuron_grads(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: Module,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     gradient_neuron_selector: None = None,
     grad_enabled: bool = False,
     device_ids: Union[None, List[int]] = None,
@@ -475,7 +470,7 @@ def _forward_layer_eval_with_neuron_grads(
     forward_fn: Callable,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     layer: ModuleOrModuleList,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
     gradient_neuron_selector: Union[
         None, int, Tuple[Union[int, slice], ...], Callable
@@ -549,8 +544,7 @@ def compute_layer_gradients_and_eval(
     layer: Module,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     target_ind: TargetType = None,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     *,
     # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
     gradient_neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
@@ -571,7 +565,7 @@ def compute_layer_gradients_and_eval(
     layer: List[Module],
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     target_ind: TargetType = None,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     gradient_neuron_selector: None = None,
     device_ids: Union[None, List[int]] = None,
     attribute_to_layer_input: bool = False,
@@ -590,7 +584,7 @@ def compute_layer_gradients_and_eval(
     layer: Module,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     target_ind: TargetType = None,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     gradient_neuron_selector: None = None,
     device_ids: Union[None, List[int]] = None,
     attribute_to_layer_input: bool = False,
@@ -606,7 +600,7 @@ def compute_layer_gradients_and_eval(
     layer: ModuleOrModuleList,
     inputs: Union[Tensor, Tuple[Tensor, ...]],
     target_ind: TargetType = None,
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
     gradient_neuron_selector: Union[
         None, int, Tuple[Union[int, slice], ...], Callable
@@ -792,8 +786,7 @@ def grad_fn(
         forward_fn: Callable,
         inputs: TensorOrTupleOfTensorsGeneric,
         target_ind: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
     ) -> Tuple[Tensor, ...]:
         _, grads = _forward_layer_eval_with_neuron_grads(
             forward_fn,
diff --git a/captum/attr/_core/deep_lift.py b/captum/attr/_core/deep_lift.py
index 8bd9dbbfe..f909bf79c 100644
--- a/captum/attr/_core/deep_lift.py
+++ b/captum/attr/_core/deep_lift.py
@@ -3,7 +3,7 @@
 # pyre-strict
 import typing
 import warnings
-from typing import Callable, cast, Dict, List, Literal, Tuple, Type, Union
+from typing import Callable, cast, Dict, List, Literal, Optional, Tuple, Type, Union
 
 import torch
 import torch.nn as nn
@@ -117,7 +117,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         *,
         return_convergence_delta: Literal[True],
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
@@ -129,7 +129,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         return_convergence_delta: Literal[False] = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
     ) -> TensorOrTupleOfTensorsGeneric: ...
@@ -140,7 +140,7 @@ def attribute(  # type: ignore
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         return_convergence_delta: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
     ) -> Union[
@@ -370,7 +370,7 @@ def _construct_forward_func(
         forward_func: Callable[..., Tensor],
         inputs: Tuple[Tuple[Tensor, ...], Tuple[Tensor, ...]],
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
     ) -> Callable[[], Tensor]:
         def forward_fn() -> Tensor:
             model_out = cast(
@@ -604,7 +604,7 @@ def attribute(
             TensorOrTupleOfTensorsGeneric, Callable[..., TensorOrTupleOfTensorsGeneric]
         ],
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         *,
         return_convergence_delta: Literal[True],
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
@@ -618,7 +618,7 @@ def attribute(
             TensorOrTupleOfTensorsGeneric, Callable[..., TensorOrTupleOfTensorsGeneric]
         ],
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         return_convergence_delta: Literal[False] = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
     ) -> TensorOrTupleOfTensorsGeneric: ...
@@ -631,7 +631,7 @@ def attribute(  # type: ignore
             TensorOrTupleOfTensorsGeneric, Callable[..., TensorOrTupleOfTensorsGeneric]
         ],
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         return_convergence_delta: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
     ) -> Union[
@@ -840,7 +840,7 @@ def _expand_inputs_baselines_targets(
         baselines: Tuple[Tensor, ...],
         inputs: Tuple[Tensor, ...],
         target: TargetType,
-        additional_forward_args: object,
+        additional_forward_args: Optional[Tuple[object, ...]],
     ) -> Tuple[Tuple[Tensor, ...], Tuple[Tensor, ...], TargetType, object]:
         inp_bsz = inputs[0].shape[0]
         base_bsz = baselines[0].shape[0]
diff --git a/captum/attr/_core/feature_ablation.py b/captum/attr/_core/feature_ablation.py
index abdb7e53f..03e66ed3e 100644
--- a/captum/attr/_core/feature_ablation.py
+++ b/captum/attr/_core/feature_ablation.py
@@ -75,7 +75,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
@@ -408,7 +408,7 @@ def attribute_future(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
@@ -655,7 +655,7 @@ def _ith_input_ablation_generator(
         self,
         i: int,
         inputs: TensorOrTupleOfTensorsGeneric,
-        additional_args: object,
+        additional_args: Optional[Tuple[object, ...]],
         target: TargetType,
         baselines: BaselineType,
         input_mask: Union[None, Tensor, Tuple[Tensor, ...]],
diff --git a/captum/attr/_core/feature_permutation.py b/captum/attr/_core/feature_permutation.py
index 19287b6ec..79c519602 100644
--- a/captum/attr/_core/feature_permutation.py
+++ b/captum/attr/_core/feature_permutation.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, Tuple, Union
+from typing import Any, Callable, Optional, Tuple, Union
 
 import torch
 from captum._utils.typing import TargetType, TensorOrTupleOfTensorsGeneric
@@ -99,7 +99,7 @@ def attribute(  # type: ignore
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, TensorOrTupleOfTensorsGeneric] = None,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
@@ -280,7 +280,7 @@ def attribute_future(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, TensorOrTupleOfTensorsGeneric] = None,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
diff --git a/captum/attr/_core/gradient_shap.py b/captum/attr/_core/gradient_shap.py
index 51328c7cd..9cf9e85a4 100644
--- a/captum/attr/_core/gradient_shap.py
+++ b/captum/attr/_core/gradient_shap.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 import typing
-from typing import Callable, Literal, Tuple, Union
+from typing import Callable, Literal, Optional, Tuple, Union
 
 import numpy as np
 import torch
@@ -91,7 +91,7 @@ def attribute(
         n_samples: int = 5,
         stdevs: Union[float, Tuple[float, ...]] = 0.0,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
     ) -> Tuple[TensorOrTupleOfTensorsGeneric, Tensor]: ...
@@ -106,7 +106,7 @@ def attribute(
         n_samples: int = 5,
         stdevs: Union[float, Tuple[float, ...]] = 0.0,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
     ) -> TensorOrTupleOfTensorsGeneric: ...
 
@@ -122,7 +122,7 @@ def attribute(
         n_samples: int = 5,
         stdevs: Union[float, Tuple[float, ...]] = 0.0,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
     ) -> Union[
         TensorOrTupleOfTensorsGeneric, Tuple[TensorOrTupleOfTensorsGeneric, Tensor]
@@ -336,7 +336,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
     ) -> Tuple[TensorOrTupleOfTensorsGeneric, Tensor]: ...
@@ -347,7 +347,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
     ) -> TensorOrTupleOfTensorsGeneric: ...
 
@@ -357,7 +357,7 @@ def attribute(  # type: ignore
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
     ) -> Union[
         TensorOrTupleOfTensorsGeneric, Tuple[TensorOrTupleOfTensorsGeneric, Tensor]
diff --git a/captum/attr/_core/guided_backprop_deconvnet.py b/captum/attr/_core/guided_backprop_deconvnet.py
index 5dc444e45..35b7f1993 100644
--- a/captum/attr/_core/guided_backprop_deconvnet.py
+++ b/captum/attr/_core/guided_backprop_deconvnet.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 import warnings
-from typing import Callable, List, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 import torch
 import torch.nn.functional as F
@@ -45,7 +45,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
         Computes attribution by overriding relu gradients. Based on constructor
@@ -146,7 +146,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
         Args:
@@ -255,7 +255,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
         Args:
diff --git a/captum/attr/_core/guided_grad_cam.py b/captum/attr/_core/guided_grad_cam.py
index d4d197849..9f89f387d 100644
--- a/captum/attr/_core/guided_grad_cam.py
+++ b/captum/attr/_core/guided_grad_cam.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 import warnings
-from typing import List, Union
+from typing import List, Optional, Union
 
 import torch
 from captum._utils.common import _format_output, _format_tensor_into_tuples, _is_tuple
@@ -72,7 +72,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         interpolate_mode: str = "nearest",
         attribute_to_layer_input: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric:
diff --git a/captum/attr/_core/input_x_gradient.py b/captum/attr/_core/input_x_gradient.py
index bfaa75def..8686a0557 100644
--- a/captum/attr/_core/input_x_gradient.py
+++ b/captum/attr/_core/input_x_gradient.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Callable
+from typing import Callable, Optional
 
 from captum._utils.common import _format_output, _format_tensor_into_tuples, _is_tuple
 from captum._utils.gradient import (
@@ -35,7 +35,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
         Args:
diff --git a/captum/attr/_core/integrated_gradients.py b/captum/attr/_core/integrated_gradients.py
index 1abbcc69f..825c2cae6 100644
--- a/captum/attr/_core/integrated_gradients.py
+++ b/captum/attr/_core/integrated_gradients.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 import typing
-from typing import Any, Callable, List, Literal, Tuple, Union
+from typing import Callable, List, Literal, Optional, Tuple, Union
 
 import torch
 from captum._utils.common import (
@@ -79,7 +79,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -95,8 +95,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -109,7 +108,7 @@ def attribute(  # type: ignore
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -328,7 +327,7 @@ def _attribute(
         inputs: Tuple[Tensor, ...],
         baselines: Tuple[Union[Tensor, int, float], ...],
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         step_sizes_and_alphas: Union[None, Tuple[List[float], List[float]]] = None,
diff --git a/captum/attr/_core/kernel_shap.py b/captum/attr/_core/kernel_shap.py
index 89d22990d..6fdbfcb9b 100644
--- a/captum/attr/_core/kernel_shap.py
+++ b/captum/attr/_core/kernel_shap.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 
-from typing import Callable, cast, Generator, Tuple, Union
+from typing import Callable, cast, Generator, Optional, Tuple, Union
 
 import torch
 from captum._utils.models.linear_model import SkLearnLinearRegression
@@ -49,7 +49,7 @@ def attribute(  # type: ignore
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         n_samples: int = 25,
         perturbations_per_eval: int = 1,
diff --git a/captum/attr/_core/layer/grad_cam.py b/captum/attr/_core/layer/grad_cam.py
index 01c14a405..eed639760 100644
--- a/captum/attr/_core/layer/grad_cam.py
+++ b/captum/attr/_core/layer/grad_cam.py
@@ -82,8 +82,7 @@ def attribute(
         self,
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_layer_input: bool = False,
         relu_attributions: bool = False,
         attr_dim_summation: bool = True,
diff --git a/captum/attr/_core/layer/internal_influence.py b/captum/attr/_core/layer/internal_influence.py
index 548d8d228..47b69ffb2 100644
--- a/captum/attr/_core/layer/internal_influence.py
+++ b/captum/attr/_core/layer/internal_influence.py
@@ -72,8 +72,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -254,8 +253,7 @@ def _attribute(
         inputs: Tuple[Tensor, ...],
         baselines: Tuple[Union[Tensor, int, float], ...],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         attribute_to_layer_input: bool = False,
diff --git a/captum/attr/_core/layer/layer_activation.py b/captum/attr/_core/layer/layer_activation.py
index 9ef5c3866..076323a27 100644
--- a/captum/attr/_core/layer/layer_activation.py
+++ b/captum/attr/_core/layer/layer_activation.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, cast, List, Tuple, Union
+from typing import Callable, cast, List, Optional, Tuple, Union
 
 import torch
 from captum._utils.common import _format_output
@@ -51,8 +51,7 @@ def __init__(
     def attribute(
         self,
         inputs: Union[Tensor, Tuple[Tensor, ...]],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_layer_input: bool = False,
     ) -> Union[Tensor, Tuple[Tensor, ...], List[Union[Tensor, Tuple[Tensor, ...]]]]:
         r"""
diff --git a/captum/attr/_core/layer/layer_conductance.py b/captum/attr/_core/layer/layer_conductance.py
index 54ec6fdb2..1f1a5f467 100644
--- a/captum/attr/_core/layer/layer_conductance.py
+++ b/captum/attr/_core/layer/layer_conductance.py
@@ -80,8 +80,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -99,7 +98,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -118,7 +117,7 @@ def attribute(
             None, int, float, Tensor, Tuple[Union[int, float, Tensor], ...]
         ] = None,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -333,8 +332,7 @@ def _attribute(
         inputs: Tuple[Tensor, ...],
         baselines: Tuple[Union[Tensor, int, float], ...],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         attribute_to_layer_input: bool = False,
diff --git a/captum/attr/_core/layer/layer_deep_lift.py b/captum/attr/_core/layer/layer_deep_lift.py
index 85d81cd5e..a126971cf 100644
--- a/captum/attr/_core/layer/layer_deep_lift.py
+++ b/captum/attr/_core/layer/layer_deep_lift.py
@@ -101,8 +101,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
         attribute_to_layer_input: bool = False,
@@ -116,8 +115,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
         attribute_to_layer_input: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
@@ -132,7 +130,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
         attribute_to_layer_input: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
@@ -445,8 +443,7 @@ def attribute(
             Tensor, Tuple[Tensor, ...], Callable[..., Union[Tensor, Tuple[Tensor, ...]]]
         ],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         *,
         return_convergence_delta: Literal[True],
         attribute_to_layer_input: bool = False,
@@ -463,8 +460,7 @@ def attribute(
             Tensor, Tuple[Tensor, ...], Callable[..., Union[Tensor, Tuple[Tensor, ...]]]
         ],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         return_convergence_delta: Literal[False] = False,
         attribute_to_layer_input: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
@@ -480,7 +476,7 @@ def attribute(
             Tensor, Tuple[Tensor, ...], Callable[..., Union[Tensor, Tuple[Tensor, ...]]]
         ],
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         return_convergence_delta: bool = False,
         attribute_to_layer_input: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
diff --git a/captum/attr/_core/layer/layer_feature_ablation.py b/captum/attr/_core/layer/layer_feature_ablation.py
index 35233cbc8..c0297d954 100644
--- a/captum/attr/_core/layer/layer_feature_ablation.py
+++ b/captum/attr/_core/layer/layer_feature_ablation.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, cast, Dict, List, Tuple, Type, Union
+from typing import Any, Callable, cast, Dict, List, Optional, Tuple, Type, Union
 
 import torch
 from captum._utils.common import (
@@ -69,7 +69,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         layer_baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         layer_mask: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         attribute_to_layer_input: bool = False,
         perturbations_per_eval: int = 1,
diff --git a/captum/attr/_core/layer/layer_feature_permutation.py b/captum/attr/_core/layer/layer_feature_permutation.py
index aee338fe6..8db7b965d 100644
--- a/captum/attr/_core/layer/layer_feature_permutation.py
+++ b/captum/attr/_core/layer/layer_feature_permutation.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, cast, Dict, List, Tuple, Type, Union
+from typing import Any, Callable, cast, Dict, List, Optional, Tuple, Type, Union
 
 import torch
 from captum._utils.common import (
@@ -62,7 +62,7 @@ def attribute(
         self,
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         layer_mask: Union[None, TensorOrTupleOfTensorsGeneric] = None,
         perturbations_per_eval: int = 1,
     ) -> Union[Tensor, Tuple[Tensor, ...]]:
diff --git a/captum/attr/_core/layer/layer_gradient_shap.py b/captum/attr/_core/layer/layer_gradient_shap.py
index dcfe109fa..c9987eb00 100644
--- a/captum/attr/_core/layer/layer_gradient_shap.py
+++ b/captum/attr/_core/layer/layer_gradient_shap.py
@@ -114,8 +114,7 @@ def attribute(
         n_samples: int = 5,
         stdevs: Union[float, Tuple[float, ...]] = 0.0,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
         attribute_to_layer_input: bool = False,
@@ -132,7 +131,7 @@ def attribute(
         n_samples: int = 5,
         stdevs: Union[float, Tuple[float, ...]] = 0.0,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
         attribute_to_layer_input: bool = False,
     ) -> Union[Tensor, Tuple[Tensor, ...]]: ...
@@ -146,7 +145,7 @@ def attribute(
         n_samples: int = 5,
         stdevs: Union[float, Tuple[float, ...]] = 0.0,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
         attribute_to_layer_input: bool = False,
     ) -> Union[
@@ -392,8 +391,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
         attribute_to_layer_input: bool = False,
@@ -406,8 +404,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
         attribute_to_layer_input: bool = False,
         grad_kwargs: Optional[Dict[str, Any]] = None,
@@ -419,7 +416,7 @@ def attribute(  # type: ignore
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
         attribute_to_layer_input: bool = False,
         grad_kwargs: Optional[Dict[str, Any]] = None,
diff --git a/captum/attr/_core/layer/layer_gradient_x_activation.py b/captum/attr/_core/layer/layer_gradient_x_activation.py
index ebbb83655..c828a262e 100644
--- a/captum/attr/_core/layer/layer_gradient_x_activation.py
+++ b/captum/attr/_core/layer/layer_gradient_x_activation.py
@@ -77,8 +77,7 @@ def attribute(
         self,
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_layer_input: bool = False,
         grad_kwargs: Optional[Dict[str, Any]] = None,
     ) -> Union[Tensor, Tuple[Tensor, ...], List[Union[Tensor, Tuple[Tensor, ...]]]]:
diff --git a/captum/attr/_core/layer/layer_integrated_gradients.py b/captum/attr/_core/layer/layer_integrated_gradients.py
index 5c56e858d..01067ca0d 100644
--- a/captum/attr/_core/layer/layer_integrated_gradients.py
+++ b/captum/attr/_core/layer/layer_integrated_gradients.py
@@ -3,7 +3,7 @@
 # pyre-strict
 import functools
 import warnings
-from typing import Any, Callable, cast, List, Literal, overload, Tuple, Union
+from typing import Callable, cast, List, Literal, Optional, overload, Tuple, Union
 
 import torch
 from captum._utils.common import (
@@ -120,8 +120,7 @@ def _gradient_func(
             forward_fn: Callable,
             inputs: Union[Tensor, Tuple[Tensor, ...]],
             target_ind: TargetType = None,
-            # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-            additional_forward_args: Any = None,
+            additional_forward_args: Optional[object] = None,
         ) -> Tuple[Tensor, ...]:
             if self.device_ids is None or len(self.device_ids) == 0:
                 scattered_inputs = (inputs,)
@@ -232,8 +231,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType,
         target: TargetType,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any,
+        additional_forward_args: Optional[object],
         n_steps: int,
         method: str,
         internal_batch_size: Union[None, int],
@@ -247,8 +245,7 @@ def attribute(  # type: ignore
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType,
         target: TargetType,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any,
+        additional_forward_args: Optional[object],
         n_steps: int,
         method: str,
         internal_batch_size: Union[None, int],
@@ -267,7 +264,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
@@ -289,7 +286,7 @@ def attribute(
         inputs: Union[Tensor, Tuple[Tensor, ...]],
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
diff --git a/captum/attr/_core/layer/layer_lrp.py b/captum/attr/_core/layer/layer_lrp.py
index cd774d855..ba6a73d70 100644
--- a/captum/attr/_core/layer/layer_lrp.py
+++ b/captum/attr/_core/layer/layer_lrp.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 import typing
-from typing import Any, cast, List, Literal, Tuple, Union
+from typing import Any, cast, List, Literal, Optional, Tuple, Union
 
 from captum._utils.common import (
     _format_tensor_into_tuples,
@@ -67,8 +67,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
         attribute_to_layer_input: bool = False,
@@ -83,8 +82,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
         attribute_to_layer_input: bool = False,
         verbose: bool = False,
@@ -94,7 +92,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
         attribute_to_layer_input: bool = False,
         verbose: bool = False,
diff --git a/captum/attr/_core/lime.py b/captum/attr/_core/lime.py
index 152bae2c3..8b5a6f86b 100644
--- a/captum/attr/_core/lime.py
+++ b/captum/attr/_core/lime.py
@@ -244,7 +244,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         n_samples: int = 50,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
@@ -876,7 +876,7 @@ def attribute(  # type: ignore
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         n_samples: int = 25,
         perturbations_per_eval: int = 1,
@@ -1121,7 +1121,7 @@ def _attribute_kwargs(  # type: ignore
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         n_samples: int = 25,
         perturbations_per_eval: int = 1,
diff --git a/captum/attr/_core/lrp.py b/captum/attr/_core/lrp.py
index d08b7b4de..c2c0dac74 100644
--- a/captum/attr/_core/lrp.py
+++ b/captum/attr/_core/lrp.py
@@ -4,7 +4,7 @@
 
 import typing
 from collections import defaultdict
-from typing import Any, Callable, cast, Dict, List, Literal, Tuple, Union
+from typing import Any, Callable, cast, Dict, List, Literal, Optional, Tuple, Union
 
 import torch.nn as nn
 from captum._utils.common import (
@@ -72,7 +72,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         *,
         return_convergence_delta: Literal[True],
         verbose: bool = False,
@@ -83,7 +83,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: Literal[False] = False,
         verbose: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric: ...
@@ -95,7 +95,7 @@ def attribute(
         self,
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         return_convergence_delta: bool = False,
         verbose: bool = False,
     ) -> Union[
@@ -367,7 +367,7 @@ def _compute_output_and_change_weights(
         self,
         inputs: Tuple[Tensor, ...],
         target: TargetType,
-        additional_forward_args: object,
+        additional_forward_args: Optional[object],
     ) -> Tensor:
         try:
             self._register_weight_hooks()
diff --git a/captum/attr/_core/neuron/neuron_conductance.py b/captum/attr/_core/neuron/neuron_conductance.py
index fcb8bafb5..359736d0c 100644
--- a/captum/attr/_core/neuron/neuron_conductance.py
+++ b/captum/attr/_core/neuron/neuron_conductance.py
@@ -98,8 +98,7 @@ def attribute(
         neuron_selector: Union[int, Tuple[int, ...], Callable],
         baselines: BaselineType = None,
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "riemann_trapezoid",
         internal_batch_size: Union[None, int] = None,
@@ -339,8 +338,7 @@ def _attribute(
         neuron_selector: Union[int, Tuple[int, ...], Callable],
         baselines: Tuple[Union[Tensor, int, float], ...],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "riemann_trapezoid",
         attribute_to_neuron_input: bool = False,
diff --git a/captum/attr/_core/neuron/neuron_deep_lift.py b/captum/attr/_core/neuron/neuron_deep_lift.py
index da7008372..f4648b43b 100644
--- a/captum/attr/_core/neuron/neuron_deep_lift.py
+++ b/captum/attr/_core/neuron/neuron_deep_lift.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, cast, Tuple, Union
+from typing import Callable, cast, Optional, Tuple, Union
 
 from captum._utils.gradient import construct_neuron_grad_fn
 from captum._utils.typing import BaselineType, TensorOrTupleOfTensorsGeneric
@@ -82,8 +82,7 @@ def attribute(
         # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
         neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
         baselines: BaselineType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_neuron_input: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
@@ -315,8 +314,7 @@ def attribute(
         baselines: Union[
             TensorOrTupleOfTensorsGeneric, Callable[..., TensorOrTupleOfTensorsGeneric]
         ],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_neuron_input: bool = False,
         custom_attribution_func: Union[None, Callable[..., Tuple[Tensor, ...]]] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
diff --git a/captum/attr/_core/neuron/neuron_feature_ablation.py b/captum/attr/_core/neuron/neuron_feature_ablation.py
index dfdd9833e..c72cf806a 100644
--- a/captum/attr/_core/neuron/neuron_feature_ablation.py
+++ b/captum/attr/_core/neuron/neuron_feature_ablation.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, cast, List, Tuple, Union
+from typing import Any, Callable, cast, List, Optional, Tuple, Union
 
 import torch
 from captum._utils.common import _verify_select_neuron
@@ -64,8 +64,7 @@ def attribute(
         # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
         neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
         baselines: BaselineType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, TensorOrTupleOfTensorsGeneric] = None,
         attribute_to_neuron_input: bool = False,
         perturbations_per_eval: int = 1,
diff --git a/captum/attr/_core/neuron/neuron_gradient.py b/captum/attr/_core/neuron/neuron_gradient.py
index fef1f2c19..0e74382d3 100644
--- a/captum/attr/_core/neuron/neuron_gradient.py
+++ b/captum/attr/_core/neuron/neuron_gradient.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, List, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 from captum._utils.common import (
     _format_additional_forward_args,
@@ -62,8 +62,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
         neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_neuron_input: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
diff --git a/captum/attr/_core/neuron/neuron_gradient_shap.py b/captum/attr/_core/neuron/neuron_gradient_shap.py
index 897dea197..18b650723 100644
--- a/captum/attr/_core/neuron/neuron_gradient_shap.py
+++ b/captum/attr/_core/neuron/neuron_gradient_shap.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, List, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 from captum._utils.gradient import construct_neuron_grad_fn
 from captum._utils.typing import TensorOrTupleOfTensorsGeneric
@@ -104,8 +104,7 @@ def attribute(
         ],
         n_samples: int = 5,
         stdevs: float = 0.0,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_neuron_input: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
diff --git a/captum/attr/_core/neuron/neuron_guided_backprop_deconvnet.py b/captum/attr/_core/neuron/neuron_guided_backprop_deconvnet.py
index d95edf2e3..03f3e1418 100644
--- a/captum/attr/_core/neuron/neuron_guided_backprop_deconvnet.py
+++ b/captum/attr/_core/neuron/neuron_guided_backprop_deconvnet.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, List, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 from captum._utils.gradient import construct_neuron_grad_fn
 from captum._utils.typing import TensorOrTupleOfTensorsGeneric
@@ -62,8 +62,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
         neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_neuron_input: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
@@ -218,8 +217,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
         neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         attribute_to_neuron_input: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
diff --git a/captum/attr/_core/neuron/neuron_integrated_gradients.py b/captum/attr/_core/neuron/neuron_integrated_gradients.py
index 3bef9caa0..8e56221d7 100644
--- a/captum/attr/_core/neuron/neuron_integrated_gradients.py
+++ b/captum/attr/_core/neuron/neuron_integrated_gradients.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, List, Tuple, Union
+from typing import Callable, List, Optional, Tuple, Union
 
 from captum._utils.gradient import construct_neuron_grad_fn
 from captum._utils.typing import TensorOrTupleOfTensorsGeneric
@@ -79,8 +79,7 @@ def attribute(
         # pyre-fixme[24]: Generic type `Callable` expects 2 type parameters.
         neuron_selector: Union[int, Tuple[Union[int, slice], ...], Callable],
         baselines: Union[None, Tensor, Tuple[Tensor, ...]] = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         n_steps: int = 50,
         method: str = "gausslegendre",
         internal_batch_size: Union[None, int] = None,
diff --git a/captum/attr/_core/occlusion.py b/captum/attr/_core/occlusion.py
index 62ac38e84..8361d1058 100644
--- a/captum/attr/_core/occlusion.py
+++ b/captum/attr/_core/occlusion.py
@@ -57,7 +57,7 @@ def attribute(  # type: ignore
         ] = None,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
     ) -> TensorOrTupleOfTensorsGeneric:
diff --git a/captum/attr/_core/saliency.py b/captum/attr/_core/saliency.py
index 8698099db..f4dce70cd 100644
--- a/captum/attr/_core/saliency.py
+++ b/captum/attr/_core/saliency.py
@@ -2,7 +2,7 @@
 
 # pyre-strict
 
-from typing import Callable
+from typing import Callable, Optional
 
 import torch
 from captum._utils.common import _format_output, _format_tensor_into_tuples, _is_tuple
@@ -41,7 +41,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         target: TargetType = None,
         abs: bool = True,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
         r"""
         Args:
diff --git a/captum/attr/_core/shapley_value.py b/captum/attr/_core/shapley_value.py
index 8f8d79137..83f1811ae 100644
--- a/captum/attr/_core/shapley_value.py
+++ b/captum/attr/_core/shapley_value.py
@@ -5,7 +5,7 @@
 import itertools
 import math
 import warnings
-from typing import Callable, cast, Iterable, Sequence, Tuple, Union
+from typing import Callable, cast, Iterable, Optional, Sequence, Tuple, Union
 
 import torch
 from captum._utils.common import (
@@ -108,7 +108,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[Tuple[object, ...]] = None,
         feature_mask: Union[None, TensorOrTupleOfTensorsGeneric] = None,
         n_samples: int = 25,
         perturbations_per_eval: int = 1,
@@ -464,7 +464,7 @@ def attribute_future(self) -> Callable:
     def _perturbation_generator(
         self,
         inputs: Tuple[Tensor, ...],
-        additional_args: object,
+        additional_args: Optional[Tuple[object, ...]],
         target: TargetType,
         baselines: Tuple[Tensor, ...],
         input_masks: TensorOrTupleOfTensorsGeneric,
@@ -627,7 +627,7 @@ def attribute(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: TargetType = None,
-        additional_forward_args: object = None,
+        additional_forward_args: Optional[object] = None,
         feature_mask: Union[None, TensorOrTupleOfTensorsGeneric] = None,
         perturbations_per_eval: int = 1,
         show_progress: bool = False,
diff --git a/captum/attr/_utils/attribution.py b/captum/attr/_utils/attribution.py
index 9cb9b297b..04f0b1d24 100644
--- a/captum/attr/_utils/attribution.py
+++ b/captum/attr/_utils/attribution.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-strict
-from typing import Any, Callable, cast, Generic, List, Tuple, Type, Union
+from typing import Callable, cast, Generic, List, Optional, Tuple, Type, Union
 
 import torch
 import torch.nn.functional as F
@@ -209,8 +209,7 @@ def compute_convergence_delta(
         ],
         end_point: Union[Tensor, Tuple[Tensor, ...]],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
     ) -> Tensor:
         r"""
         Here we provide a specific implementation for `compute_convergence_delta`
diff --git a/captum/attr/_utils/batching.py b/captum/attr/_utils/batching.py
index 96e1e20a4..f7bce61ec 100644
--- a/captum/attr/_utils/batching.py
+++ b/captum/attr/_utils/batching.py
@@ -3,7 +3,7 @@
 # pyre-strict
 import typing
 import warnings
-from typing import Any, Callable, Iterator, Tuple, Union
+from typing import Any, Callable, Iterator, Optional, Tuple, Union
 
 import torch
 from captum._utils.common import (
@@ -139,8 +139,7 @@ def _tuple_splice_range(
 # pyre-fixme[3]: Return annotation cannot contain `Any`.
 def _batched_generator(
     inputs: TensorOrTupleOfTensorsGeneric,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     target_ind: TargetType = None,
     internal_batch_size: Union[None, int] = None,
 ) -> Iterator[Tuple[Tuple[Tensor, ...], Any, TargetType]]:
@@ -181,6 +180,9 @@ def _batched_generator(
                 )
             # pyre-fixme[7]: Expected `Iterator[Tuple[typing.Tuple[Tensor, ...], typi...
             yield inputs_splice, _tuple_splice_range(
+                # pyre-fixme[6]: In call `_tuple_splice_range`, for 1st positional
+                # argument, expected `None` but got
+                # `Optional[typing.Tuple[typing.Any, ...]]`
                 additional_forward_args,
                 current_total,
                 current_total + internal_batch_size,
@@ -195,8 +197,7 @@ def _batched_generator(
 def _batched_operator(
     operator: Callable[..., TupleOrTensorOrBoolGeneric],
     inputs: TensorOrTupleOfTensorsGeneric,
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any = None,
+    additional_forward_args: Optional[object] = None,
     target_ind: TargetType = None,
     internal_batch_size: Union[None, int] = None,
     **kwargs: Any,
diff --git a/captum/attr/_utils/common.py b/captum/attr/_utils/common.py
index 09889cd52..2c03acc03 100644
--- a/captum/attr/_utils/common.py
+++ b/captum/attr/_utils/common.py
@@ -3,7 +3,7 @@
 # pyre-strict
 import typing
 from inspect import signature
-from typing import Any, Callable, List, Literal, Tuple, TYPE_CHECKING, Union
+from typing import Callable, List, Literal, Optional, Tuple, TYPE_CHECKING, Union
 
 import torch
 from captum._utils.common import (
@@ -207,8 +207,7 @@ def _compute_conv_delta_and_format_attrs(
     attributions: Tuple[Tensor, ...],
     start_point: Union[int, float, Tensor, Tuple[Union[int, float, Tensor], ...]],
     end_point: Union[Tensor, Tuple[Tensor, ...]],
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any,
+    additional_forward_args: Optional[object],
     target: TargetType,
     is_inputs_tuple: Literal[True],
 ) -> Union[Tuple[Tensor, ...], Tuple[Tuple[Tensor, ...], Tensor]]: ...
@@ -221,8 +220,7 @@ def _compute_conv_delta_and_format_attrs(
     attributions: Tuple[Tensor, ...],
     start_point: Union[int, float, Tensor, Tuple[Union[int, float, Tensor], ...]],
     end_point: Union[Tensor, Tuple[Tensor, ...]],
-    # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-    additional_forward_args: Any,
+    additional_forward_args: Optional[object],
     target: TargetType,
     is_inputs_tuple: Literal[False] = False,
 ) -> Union[Tensor, Tuple[Tensor, Tensor]]: ...
@@ -235,7 +233,7 @@ def _compute_conv_delta_and_format_attrs(
     attributions: Tuple[Tensor, ...],
     start_point: Union[int, float, Tensor, Tuple[Union[int, float, Tensor], ...]],
     end_point: Union[Tensor, Tuple[Tensor, ...]],
-    additional_forward_args: Any,
+    additional_forward_args: Optional[object],
     target: TargetType,
     is_inputs_tuple: bool = False,
 ) -> Union[
@@ -251,7 +249,7 @@ def _compute_conv_delta_and_format_attrs(
     attributions: Tuple[Tensor, ...],
     start_point: Union[int, float, Tensor, Tuple[Union[int, float, Tensor], ...]],
     end_point: Union[Tensor, Tuple[Tensor, ...]],
-    additional_forward_args: Any,
+    additional_forward_args: Optional[object],
     target: TargetType,
     is_inputs_tuple: bool = False,
 ) -> Union[
diff --git a/captum/concept/_core/tcav.py b/captum/concept/_core/tcav.py
index a81c50079..daea9f939 100644
--- a/captum/concept/_core/tcav.py
+++ b/captum/concept/_core/tcav.py
@@ -589,8 +589,7 @@ def interpret(
         inputs: TensorOrTupleOfTensorsGeneric,
         experimental_sets: List[List[Concept]],
         target: TargetType = None,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         processes: Optional[int] = None,
         **kwargs: Any,
     ) -> Dict[str, Dict[str, Dict[str, Tensor]]]:
diff --git a/captum/metrics/_core/infidelity.py b/captum/metrics/_core/infidelity.py
index 83e770592..c4c4bd061 100644
--- a/captum/metrics/_core/infidelity.py
+++ b/captum/metrics/_core/infidelity.py
@@ -147,7 +147,7 @@ def infidelity(
     inputs: TensorOrTupleOfTensorsGeneric,
     attributions: TensorOrTupleOfTensorsGeneric,
     baselines: BaselineType = None,
-    additional_forward_args: object = None,
+    additional_forward_args: Optional[object] = None,
     target: TargetType = None,
     n_perturb_samples: int = 10,
     max_examples_per_batch: Optional[int] = None,
@@ -571,7 +571,7 @@ def _make_next_infidelity_tensors_func(
     inputs: Tuple[Tensor, ...],
     baselines: BaselineTupleType,
     attributions: Tuple[Tensor, ...],
-    additional_forward_args: object = None,
+    additional_forward_args: Optional[Tuple[object, ...]] = None,
     target: TargetType = None,
     normalize: bool = False,
 ) -> Callable[[int], Union[Tuple[Tensor], Tuple[Tensor, Tensor, Tensor]]]:
diff --git a/captum/robust/_core/fgsm.py b/captum/robust/_core/fgsm.py
index e57deb0b5..af36e25ba 100644
--- a/captum/robust/_core/fgsm.py
+++ b/captum/robust/_core/fgsm.py
@@ -87,8 +87,7 @@ def perturb(
         epsilon: float,
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         target: Any,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         targeted: bool = False,
         mask: Optional[TensorOrTupleOfTensorsGeneric] = None,
     ) -> TensorOrTupleOfTensorsGeneric:
diff --git a/captum/robust/_core/metrics/attack_comparator.py b/captum/robust/_core/metrics/attack_comparator.py
index 649fd6283..348e2d69e 100644
--- a/captum/robust/_core/metrics/attack_comparator.py
+++ b/captum/robust/_core/metrics/attack_comparator.py
@@ -259,8 +259,7 @@ def evaluate(
         self,
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         inputs: Any,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         perturbations_per_eval: int = 1,
         # pyre-fixme[2]: Parameter must be annotated.
         **kwargs,
diff --git a/captum/robust/_core/metrics/min_param_perturbation.py b/captum/robust/_core/metrics/min_param_perturbation.py
index 1c2d32fae..afca08f1c 100644
--- a/captum/robust/_core/metrics/min_param_perturbation.py
+++ b/captum/robust/_core/metrics/min_param_perturbation.py
@@ -169,8 +169,7 @@ def _evaluate_batch(
         # pyre-fixme[24]: Generic type `list` expects 1 type parameter, use
         #  `typing.List[<element type>]` to avoid runtime subscripting errors.
         input_list: List,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any,
+        additional_forward_args: Optional[Tuple[object, ...]],
         correct_fn_kwargs: Optional[Dict[str, Any]],
         target: TargetType,
     ) -> Optional[int]:
@@ -239,8 +238,7 @@ def _linear_search(
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         preproc_input: Any,
         attack_kwargs: Optional[Dict[str, Any]],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any,
+        additional_forward_args: Optional[Tuple[object, ...]],
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         expanded_additional_args: Any,
         correct_fn_kwargs: Optional[Dict[str, Any]],
@@ -301,8 +299,7 @@ def _binary_search(
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         preproc_input: Any,
         attack_kwargs: Optional[Dict[str, Any]],
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any,
+        additional_forward_args: Optional[Tuple[object, ...]],
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         expanded_additional_args: Any,
         correct_fn_kwargs: Optional[Dict[str, Any]],
diff --git a/captum/robust/_core/pgd.py b/captum/robust/_core/pgd.py
index 76d654011..cf49c26ae 100644
--- a/captum/robust/_core/pgd.py
+++ b/captum/robust/_core/pgd.py
@@ -80,8 +80,7 @@ def perturb(
         step_num: int,
         # pyre-fixme[2]: Parameter annotation cannot be `Any`.
         target: Any,
-        # pyre-fixme[2]: Parameter annotation cannot be `Any`.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         targeted: bool = False,
         random_start: bool = False,
         norm: str = "Linf",
diff --git a/tests/attr/test_input_x_gradient.py b/tests/attr/test_input_x_gradient.py
index f06bd01c7..68b442ac5 100644
--- a/tests/attr/test_input_x_gradient.py
+++ b/tests/attr/test_input_x_gradient.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-unsafe
-from typing import Any, cast
+from typing import cast, Optional
 
 import torch
 from captum._utils.typing import TensorOrTupleOfTensorsGeneric
@@ -63,7 +63,7 @@ def _input_x_gradient_base_assert(
         model: Module,
         inputs: TensorOrTupleOfTensorsGeneric,
         expected_grads: TensorOrTupleOfTensorsGeneric,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         nt_type: str = "vanilla",
     ) -> None:
         input_x_grad = InputXGradient(model)
diff --git a/tests/attr/test_integrated_gradients_basic.py b/tests/attr/test_integrated_gradients_basic.py
index 2846e119e..378f7eb49 100644
--- a/tests/attr/test_integrated_gradients_basic.py
+++ b/tests/attr/test_integrated_gradients_basic.py
@@ -3,7 +3,7 @@
 # pyre-strict
 
 import unittest
-from typing import Any, cast, Optional, Tuple, Union
+from typing import cast, Optional, Tuple, Union
 
 import torch
 from captum._utils.common import _zeros
@@ -394,9 +394,7 @@ def _compute_attribution_and_evaluate(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: BaselineType = None,
         target: Union[None, int] = None,
-        # pyre-fixme[2]: Parameter `additional_forward_args` has type `None`
-        # but type`Any` is specified.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         type: str = "vanilla",
         approximation_method: str = "gausslegendre",
         multiply_by_inputs: bool = True,
@@ -502,9 +500,7 @@ def _compute_attribution_batch_helper_evaluate(
         inputs: TensorOrTupleOfTensorsGeneric,
         baselines: Union[None, Tensor, Tuple[Tensor, ...]] = None,
         target: Optional[int] = None,
-        # pyre-fixme[2]: Parameter `additional_forward_args` has type `None` but type
-        # `Any` is specified.
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         approximation_method: str = "gausslegendre",
     ) -> None:
         ig = IntegratedGradients(model)
diff --git a/tests/attr/test_saliency.py b/tests/attr/test_saliency.py
index 1ecf3872b..e6695eaec 100644
--- a/tests/attr/test_saliency.py
+++ b/tests/attr/test_saliency.py
@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 
 # pyre-unsafe
-from typing import Any, cast, Tuple, Union
+from typing import cast, Optional, Tuple, Union
 
 import torch
 from captum._utils.typing import TensorOrTupleOfTensorsGeneric
@@ -120,7 +120,7 @@ def _saliency_base_assert(
         model: Module,
         inputs: TensorOrTupleOfTensorsGeneric,
         expected: TensorOrTupleOfTensorsGeneric,
-        additional_forward_args: Any = None,
+        additional_forward_args: Optional[object] = None,
         nt_type: str = "vanilla",
         n_samples_batch_size=None,
     ) -> Union[Tensor, Tuple[Tensor, ...]]:
diff --git a/tests/metrics/test_sensitivity.py b/tests/metrics/test_sensitivity.py
index 9fafed5e7..14c04fba5 100644
--- a/tests/metrics/test_sensitivity.py
+++ b/tests/metrics/test_sensitivity.py
@@ -3,7 +3,7 @@
 # pyre-strict
 
 import typing
-from typing import Any, Callable, cast, List, Optional, Tuple, Union
+from typing import Callable, cast, List, Optional, Tuple, Union
 
 import torch
 from captum._utils.typing import BaselineType, TargetType, TensorOrTupleOfTensorsGeneric
@@ -293,9 +293,7 @@ def sensitivity_max_assert(
         max_examples_per_batch: Optional[int] = None,
         baselines: Optional[BaselineType] = None,
         target: Optional[TargetType] = None,
-        # pyre-fixme[2]: Parameter `additional_forward_args` has type `None`
-        # but type `Any` is specified.
-        additional_forward_args: Optional[Any] = None,
+        additional_forward_args: Optional[object] = None,
     ) -> Tensor:
         if baselines is None:
             sens = sensitivity_max(