Skip to content
forked from pytorch/captum

Model interpretability and understanding for PyTorch

License

Notifications You must be signed in to change notification settings

edward-io/captum

Repository files navigation

Captum Logo

Conda PyPI CircleCI Codecov

Captum is a model interpretability and understanding library for PyTorch. Captum means comprehension in latin and contains general purpose implementations of integrated gradients, saliency maps, smoothgrad, vargrad and others for PyTorch models. It has quick integration for models built with domain-specific libraries such as torchvision, torchtext, and others.

Captum is currently in beta and under active development!

Why Captum?

TODO: Add content here

Target Audience

TODO: Add content here

Installation

Installation Requirements

  • Python >= 3.6
  • PyTorch >= 1.2
Installing the latest release

The latest release of Captum is easily installed either via Anaconda (recommended):

conda install captum -c pytorch

or via pip:

pip install captum
Installing from latest master

If you'd like to try our bleeding edge features (and don't mind potentially running into the occasional bug here or there), you can install the latest master directly from GitHub:

pip install git+https://github.com/pytorch/captum.git

Manual / Dev install

Alternatively, you can do a manual install. For a basic install, run:

git clone https://github.com/pytorch/captum.git
cd captum
pip install -e .

To customize the installation, you can also run the following variants of the above:

  • pip install -e .[dev]: Also installs all tools necessary for development (testing, linting, docs building; see Contributing below).
  • pip install -e .[tutorials]: Also installs all packages necessary for running the tutorial notebooks.

To execute unit tests from a manual install, run:

# running a single unit test
python -m unittest -v tests.attr.test_saliency
# running all unit tests
pytest -ra

Getting Started

Captum helps to interpret and understand the predictions of PyTorch models by looking at the features that contribute to the decision that the model makes. It also helps to understand which neurons and layers are important for model predictions.

To do so, currently, it uses gradient-based model interpretability algorithms and attributes contributions to each input of the model with respect to different neurons and layers, both intermediate and final.

Let's apply some of those algorithms to a toy model that we have created for demonstration purposes. For simplicity, we will use the following simple architecture, but users are welcome to use any PyTorch model of their choice.

import numpy as np

import torch
import torch.nn as nn

from captum.attr import (
    GradientShap,
    IntegratedGradients,
    LayerConductance,
    NeuronConductance,
    NoiseTunnel,
)

class ToyModel(nn.Module):
    def __init__(self):
        super().__init__()

        self.lin1 = nn.Linear(3, 4)
        self.lin1.weight = nn.Parameter(torch.ones(4, 3))
        self.lin1.bias = nn.Parameter(torch.tensor([-10.0, 1.0, 1.0, 1.0]))
        self.relu = nn.ReLU()
        self.lin2 = nn.Linear(4, 1)
        self.lin2.weight = nn.Parameter(torch.ones(1, 4))
        self.lin2.bias = nn.Parameter(torch.tensor([-3.0]))

    def forward(self, input):
        lin1 = self.lin1(input)
        relu = self.relu(lin1)
        lin2 = self.lin2(relu)
        return lin2

Let's create an instance of our model and set it to eval mode.

model = ToyModel() model.eval()


Next, we would like to define a simple input and baseline tensors.
Baselines belong to the input space and often carry no predictive signal.
Zero tensor can serve as a baseline for some tasks.
Some interpretability algorithms such as Integrated
Gradients, Deeplift, GradientShap are designed to attribute the change between
the input and baseline to a predictive class or a value that the neural
network outputs.

We will apply model interpretability algorithms on the network
mentioned above in order to understand the importance of individual
neurons/layers and the parts of the input that play an important role in the
final prediction.

Let's fix random seeds to make computations deterministic

torch.manual_seed(123) np.random.seed(124)


Let's define our input and baseline tensors. Baselines are used in some
interpretability algorithms such as `IntegratedGradients, DeepLift,
GradientShap, NeuronConductance, LayerConductance, InternalInfluence and
NeuronIntegratedGradients`.

input = torch.rand(2, 3) baseline = torch.zeros(2, 3)

Next we will use `IntegratedGradients` algorithms to assign attribution
scores to each input feature with respect to final output.

ig = IntegratedGradients(model) attributions, delta = ig.attribute(input, baseline) print('IG Attributions: ', attributions, ' Approximation error: ', delta)

Output:

IG Attributions: tensor([[0.8883, 1.5497, 0.7550], [2.0657, 0.2219, 2.5996]]) Approximation Error: 9.5367431640625e-07

The algorithm outputs an attribution score for each input element and an
approximation error that we would like to minimize. If the approximation error
is large, we can try larger number of integral approximation steps by setting
`n_steps` to a larger value. Not all algorithms return approximation error.
Those which do, they compute it based on the completeness property of the algorithms.

Positive attribution score means that the input in that particular position positively
contributed to the final prediction and negative means the opposite.
The magnitude of the attribution score signifies the strength of the contribution.
Zero attribution score means no contribution from that particular feature.

Similarly, we can apply GradientShap, DeepLift and other attribution algorithms to the model.

gs = GradientShap(model)

We define a distribution of baselines and draw n_samples from that

distribution in order to estimate the expectations of gradients across all baselines

baseline_dist = torch.rand(100, 3) attributions, delta = gs.attribute(input, baseline_dist, n_samples=50) print('GradientShap Attributions: ', attributions, ' Approximation error: ', delta)

Output

GradientShap Attributions: tensor([[ 0.0159, -0.8478, 0.3028], [ 0.1546, -1.0068, 0.2770]]) Approximation Error: tensor(0.0462)

In order to smooth and improve the quality of the attributions we can run
`IntegratedGradients` and other attribution methods through a `NoiseTunnel`.
`NoiseTunnel` allows to use SmoothGrad, SmoothGrad_Sq and VarGrad techniques
to smoothen the attributions by aggregating them for multiple noisy
samples that were generated by adding gaussian noise.

Here is an example how we can use `NoiseTunnel` with `IntegratedGradients`.

ig = IntegratedGradients(model) nt = NoiseTunnel(ig) attributions, delta = nt.attribute(input, nt_type='smoothgrad', baselines=baseline) print('IG + SmoothGrad Attributions: ', attributions, ' Approximation error: ', delta)

Output

IG + SmoothGrad Attributions: tensor([[-1.2138, 0.6688, 0.7747], [1.3862, 0.7529, 2.2907]]) Approximation Error: 0.07243824005126953


Let's look into the internals of our network and understand which layers
and neurons are important for the predictions.
We will start with the neuron conductance. Neuron conductance helps us to identify
input features that are important for a particular neuron in a given
layer. In this case, we choose to analyze the third neuron in the first layer.

nc = NeuronConductance(model, model.lin2) attributions = nc.attribute(input, neuron_index=3, baselines=baseline) print('Neuron Attributions: ', attributions)

Output

Neuron Attributions: tensor([[0.2902, 0.5062, 0.2466], [0.6748, 0.0725, 0.8492]])


Layer conductance shows the importance of neurons for a layer and given input.
It doesn't attribute the contribution scores to the input features
but shows the importance of each neuron in selected layer.

lc = LayerConductance(model, model.lin1) attributions, delta = lc.attribute(input, baselines=baseline) print('Layer Attributions: ', attributions, ' Approximation Error: ', delta)

Outputs

Layer Attributions: tensor([[0.8883, 1.5497, 0.7550], [2.0657, 0.2219, 2.5996]], grad_fn=) Approximation error: 9.5367431640625e-07


More details on the list of supported algorithms and how to apply
Captum on different types of models can be found in our tutorials.

## Contributing
See the [CONTRIBUTING](CONTRIBUTING.md) file for how to help out.


## References

* [Axiomatic Attribution for Deep Networks, Mukund Sundararajan et al. 2017](https://arxiv.org/abs/1703.01365)
* [Did the Model Understand the Question? Pramod K. Mudrakarta, et al. 2018](https://www.aclweb.org/anthology/P18-1176)
* [Investigating the influence of noise and distractors on the interpretation of neural networks, Pieter-Jan Kindermans et al. 2016](https://arxiv.org/abs/1611.07270)
* [SmoothGrad: removing noise by adding noise, Daniel Smilkov et al. 2017](https://arxiv.org/abs/1706.03825)
* [Local Explanation Methods for Deep Neural Networks Lack Sensitivity to Parameter Values, Julius Adebayo et al. 2018](https://arxiv.org/abs/1810.03307)
* [Sanity Checks for Saliency Maps, Julius Adebayo et al. 2018](https://arxiv.org/abs/1810.03292)
* [How Important is a neuron?, Kedar Dhamdhere et al. 2018](https://arxiv.org/abs/1805.12233)
* [Learning Important Features Through Propagating Activation Differences, Avanti Shrikumar et al. 2017](https://arxiv.org/pdf/1704.02685.pdf)
* [Computationally Efficient Measures of Internal Neuron Importance, Avanti Shrikumar et al. 2018](https://arxiv.org/pdf/1807.09946.pdf)
* [A Unified Approach to Interpreting Model Predictions, Scott M. Lundberg et al. 2017](http://papers.nips.cc/paper/7062-a-unified-approach-to-interpreting-model-predictions)
* [Influence-Directed Explanations for Deep Convolutional Networks, Klas Leino et al. 2018](https://arxiv.org/pdf/1802.03788.pdf)
* [Towards better understanding of gradient-based attribution methods for deep neural networks, Marco Ancona et al. 2018](https://openreview.net/pdf?id=Sy21R9JAW)

## License
Captum is BSD licensed, as found in the [LICENSE](LICENSE) file.

About

Model interpretability and understanding for PyTorch

Resources

License

Code of conduct

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 91.5%
  • CSS 2.8%
  • JavaScript 2.6%
  • TypeScript 2.0%
  • Shell 1.0%
  • Batchfile 0.1%