Generalized Beliefs for Cooperative AI
This addresses the challenge of flexible cooperation in multi-agent AI systems, though it is incremental as it builds on existing belief and convention-aware approaches.
The paper tackles the problem of specialized conventions in self-play policies that hinder cooperation with novel partners by proposing a belief learning model that adapts to unseen conventions at test time, resulting in improved ad-hoc teamplay across various policy pools.
Self-play is a common paradigm for constructing solutions in Markov games that can yield optimal policies in collaborative settings. However, these policies often adopt highly-specialized conventions that make playing with a novel partner difficult. To address this, recent approaches rely on encoding symmetry and convention-awareness into policy training, but these require strong environmental assumptions and can complicate policy training. We therefore propose moving the learning of conventions to the belief space. Specifically, we propose a belief learning model that can maintain beliefs over rollouts of policies not seen at training time, and can thus decode and adapt to novel conventions at test time. We show how to leverage this model for both search and training of a best response over various pools of policies to greatly improve ad-hoc teamplay. We also show how our setup promotes explainability and interpretability of nuanced agent conventions.