Symmetry-Breaking Augmentations for Ad Hoc Teamwork
This work addresses the challenge of human-AI alignment in dynamic teamwork, offering a domain-specific solution for improving agent adaptability to diverse human conventions.
The paper tackles the problem of AI agents adapting to novel teammates with unforeseen strategies in collaborative settings by introducing symmetry-breaking augmentations (SBA), which increase behavioural diversity during training to improve robustness, and demonstrates that SBA outperforms previous methods in the card game Hanabi.
In dynamic collaborative settings, for artificial intelligence (AI) agents to better align with humans, they must adapt to novel teammates who utilise unforeseen strategies. While adaptation is often simple for humans, it can be challenging for AI agents. Our work introduces symmetry-breaking augmentations (SBA) as a novel approach to this challenge. By applying a symmetry-flipping operation to increase behavioural diversity among training teammates, SBA encourages agents to learn robust responses to unknown strategies, highlighting how social conventions impact human-AI alignment. We demonstrate this experimentally in two settings, showing that our approach outperforms previous ad hoc teamwork results in the challenging card game Hanabi. In addition, we propose a general metric for estimating symmetry dependency amongst a given set of policies. Our findings provide insights into how AI systems can better adapt to diverse human conventions and the core mechanics of alignment.