N$\text{A}^{\text{2}}$Q: Neural Attention Additive Model for Interpretable Multi-Agent Q-Learning

Zichuan Liu, Yuanyang Zhu, Chunlin Chen
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:22539-22558, 2023.

Abstract

Value decomposition is widely used in cooperative multi-agent reinforcement learning, however, its implicit credit assignment mechanism is not yet fully understood due to black-box networks. In this work, we study an interpretable value decomposition framework via the family of generalized additive models. We present a novel method, named Neural Attention Additive Q-learning (N$\text{A}^\text{2}$Q), providing inherent intelligibility of collaboration behavior. N$\text{A}^\text{2}$Q can explicitly factorize the optimal joint policy induced by enriching shape functions to model all possible coalition of agents into individual policies. Moreover, we construct the identity semantics to promote estimating credits together with the global state and individual value functions, where local semantic masks help us diagnose whether each agent captures the relevant-task information. Extensive experiments show that N$\text{A}^\text{2}$Q consistently achieves superior performance compared to different state-of-the-art methods on all challenging tasks, while yielding human-like interpretability.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-liu23be, title = {{N}$\text{{A}}^\text{2}${Q}: Neural Attention Additive Model for Interpretable Multi-Agent Q-Learning}, author = {Liu, Zichuan and Zhu, Yuanyang and Chen, Chunlin}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {22539--22558}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/liu23be/liu23be.pdf}, url = {https://proceedings.mlr.press/v202/liu23be.html}, abstract = {Value decomposition is widely used in cooperative multi-agent reinforcement learning, however, its implicit credit assignment mechanism is not yet fully understood due to black-box networks. In this work, we study an interpretable value decomposition framework via the family of generalized additive models. We present a novel method, named Neural Attention Additive Q-learning (N$\text{A}^\text{2}$Q), providing inherent intelligibility of collaboration behavior. N$\text{A}^\text{2}$Q can explicitly factorize the optimal joint policy induced by enriching shape functions to model all possible coalition of agents into individual policies. Moreover, we construct the identity semantics to promote estimating credits together with the global state and individual value functions, where local semantic masks help us diagnose whether each agent captures the relevant-task information. Extensive experiments show that N$\text{A}^\text{2}$Q consistently achieves superior performance compared to different state-of-the-art methods on all challenging tasks, while yielding human-like interpretability.} }
Endnote
%0 Conference Paper %T N$\text{A}^{\text{2}}$Q: Neural Attention Additive Model for Interpretable Multi-Agent Q-Learning %A Zichuan Liu %A Yuanyang Zhu %A Chunlin Chen %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-liu23be %I PMLR %P 22539--22558 %U https://proceedings.mlr.press/v202/liu23be.html %V 202 %X Value decomposition is widely used in cooperative multi-agent reinforcement learning, however, its implicit credit assignment mechanism is not yet fully understood due to black-box networks. In this work, we study an interpretable value decomposition framework via the family of generalized additive models. We present a novel method, named Neural Attention Additive Q-learning (N$\text{A}^\text{2}$Q), providing inherent intelligibility of collaboration behavior. N$\text{A}^\text{2}$Q can explicitly factorize the optimal joint policy induced by enriching shape functions to model all possible coalition of agents into individual policies. Moreover, we construct the identity semantics to promote estimating credits together with the global state and individual value functions, where local semantic masks help us diagnose whether each agent captures the relevant-task information. Extensive experiments show that N$\text{A}^\text{2}$Q consistently achieves superior performance compared to different state-of-the-art methods on all challenging tasks, while yielding human-like interpretability.
APA
Liu, Z., Zhu, Y. & Chen, C.. (2023). N$\text{A}^{\text{2}}$Q: Neural Attention Additive Model for Interpretable Multi-Agent Q-Learning. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:22539-22558 Available from https://proceedings.mlr.press/v202/liu23be.html.

Related Material