pgeon-xai 1.0

pgeon-xai is a Python package that produces explanations for opaque agents using Policy Graphs (PGs)

pgeon-xai

Tormos, A., Abalos, V., Gnatyshak, D., & Alvarez-Napagao, S. (2023, October). Policy graphs in action: explaining single- and multi-agent behaviour using predicates. In XAI in Action: Past, Present, and Future Applications.

pgeon-xai is a Python package that produces explanations for opaque agents using Policy Graphs (PGs).

A Policy Graph is a means to obtain a representation of the behavior of an opaque agent, in the form of a directed graph. Discrete states are mapped to nodes and actions to edges.

Getting started

1. Download the pgeon_xai/ folder and move it into the root directory of your project.
2. Install pgeon-xai's requirements with pip install -r PATH_TO_PGEON_REQUIREMENTS.

Example usage

Generating a Policy Graph

Given a Gymnasium environment and a discretizer, you can generate a PG to describe an opaque agent's behavior with fit().

from pgeon_xai import PolicyGraph

pg = PolicyGraph(environment, discretizer)
pg = pg.fit(agent, num_episodes=1000)

Creating and using a PG-based policy

There exist two PG-based policies. You can select one or the other with PGBasedPolicyMode.

from pgeon_xai import PGBasedPolicy, PGBasedPolicyMode

greedy_policy = PGBasedPolicy(pg, mode=PGBasedPolicyMode.GREEDY)
stochastic_policy = PGBasedPolicy(pg, mode=PGBasedPolicyMode.STOCHASTIC)

# Passing the policy an observation to get an action
obs, _ = environment.reset()
action = greedy_policy.act(obs)

More examples

You can check examples/cartpole/demo.ipynb for a complete breakdown of pgeon-xai's features.

To run the notebook yourself:

1. Download the entire repository.
2. Install pgeon-xai's requirements with pip install -r requirements.txt.
3. Install an extra dependency, rllib, with pip install "ray[rllib]".
4. Open and execute examples/cartpole/demo.ipynb.

Citation

If you use the pgeon-xai library, please cite:

@inproceedings{tormos2023policy,
  title={Policy graphs in action: explaining single-and multi-agent behaviour using predicates},
  author={Tormos, Adrian and Abalos, Victor and Gnatyshak, Dmitry and Alvarez-Napagao, Sergio},
  booktitle={XAI in Action: Past, Present, and Future Applications},
  year={2023},
  url={https://openreview.net/forum?id=QPqL9xsYOf}
}