Explainable Agency in Reinforcement Learning Agents

AAAI 2020 pp. 13724-13725

doi:10.1609/AAAI.V34I10.7134 /aaai/2020/madumal2020aaai-explainable-a/

Abstract

This thesis explores how reinforcement learning (RL) agents can provide explanations for their actions and behaviours. As humans, we build causal models to encode cause-effect relations of events and use these to explain why events happen. Taking inspiration from cognitive psychology and social science literature, I build causal explanation models and explanation dialogue models for RL agents. By mimicking human-like explanation models, these agents can provide explanations that are natural and intuitive to humans.

PDF AAAI Semantic Scholar

Cite

Text

Madumal. "Explainable Agency in Reinforcement Learning Agents." AAAI Conference on Artificial Intelligence, 2020. doi:10.1609/AAAI.V34I10.7134

Markdown

[Madumal. "Explainable Agency in Reinforcement Learning Agents." AAAI Conference on Artificial Intelligence, 2020.](https://mlanthology.org/aaai/2020/madumal2020aaai-explainable-a/) doi:10.1609/AAAI.V34I10.7134

BibTeX

@inproceedings{madumal2020aaai-explainable-a,
  title     = {{Explainable Agency in Reinforcement Learning Agents}},
  author    = {Madumal, Prashan},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2020},
  pages     = {13724-13725},
  doi       = {10.1609/AAAI.V34I10.7134},
  url       = {https://mlanthology.org/aaai/2020/madumal2020aaai-explainable-a/}
}