DiffAbs is a PyTorch implementation of multiple abstract domains that can be used in certifying or reasoning neural networks. Implemented purely using PyTorch, it is differentiable and supports GPU by default, thus amenable for safety/robustness driven training on abstract domains.
Currently, the following abstract domains are implemented:
- Vanilla interval domain;
- DeepPoly domain (https://dl.acm.org/doi/10.1145/3290354);
DeepPoly ReLU heuristics:
- A variant of the original DeepPoly domain is implemented where the
ReLU approximation is not heuristically choosing between two choices
(either picking
y = x
ory = 0
as the new upper bound). Right now it is fixed to choosingy = 0
, because there was Galois connection violation observed if this heuristic is enabled. Basically, it is observed in experiment that a smaller abstraction may unexpectedly incur larger safety distance than its containing larger abstraction.
Although it is currently tested on Mac OS X 10.15 and Ubuntu 16.04 with Python 3.7 and PyTorch 1.5, it should generalize to other platforms and older PyTorch (perhaps ≥ v1.0) smoothly.
However, Python ≤ 3.6 may be incompatible. Because type annotations
are specified everywhere and the type annotation of self class is only
supported by __future__.annotations
in Python 3.7. If using Python
3.6, this needs to use 'type string' instead.
In your virtual environment, either install directly from this repository by
git clone [email protected]:XuankangLin/DiffAbs.git
cd DiffAbs
pip install -e .
or directly install from PyPI:
pip install diffabs
Test cases for individual abstract domains are under the tests/
directory and can be run using command
pytest
Note that torchvision
is needed to run the tests for conv/maxpool layers.
The project is available open source under the terms of MIT License.