Quasi-Equivalence Discovery for Zero-Shot Emergent Communication
This addresses the challenge of generalization in multi-agent communication for real-world applications like robotics, though it is incremental as it builds on prior symmetry-based methods.
The paper tackles the problem of enabling zero-shot coordination in emergent communication by introducing the Quasi-Equivalence Discovery (QED) algorithm, which discovers symmetries to allow learned protocols to generalize to independently trained agents, achieving convergence to optimal policies in referential games with costly communication channels.
Effective communication is an important skill for enabling information exchange in multi-agent settings and emergent communication is now a vibrant field of research, with common settings involving discrete cheap-talk channels. Since, by definition, these settings involve arbitrary encoding of information, typically they do not allow for the learned protocols to generalize beyond training partners. In contrast, in this work, we present a novel problem setting and the Quasi-Equivalence Discovery (QED) algorithm that allows for zero-shot coordination (ZSC), i.e., discovering protocols that can generalize to independently trained agents. Real world problem settings often contain costly communication channels, e.g., robots have to physically move their limbs, and a non-uniform distribution over intents. We show that these two factors lead to unique optimal ZSC policies in referential games, where agents use the energy cost of the messages to communicate intent. Other-Play was recently introduced for learning optimal ZSC policies, but requires prior access to the symmetries of the problem. Instead, QED can iteratively discovers the symmetries in this setting and converges to the optimal ZSC policy.