Learning to Coordinate without Communication under Incomplete Information

TL;DR

This paper presents a method for agents to coordinate effectively in cooperative games without communication by interpreting action sequences as signals, using finite-state transducers, achieving performance close to communication-based strategies.

Contribution

It introduces a novel approach where agents learn to coordinate through action observation alone, using finite-state transducers to interpret partner behaviors.

Findings

Strategies outperform uncoordinated approaches

Performance closely matches communication-based coordination

Effective in incomplete information settings

Abstract

Achieving seamless coordination in cooperative games is a crucial challenge in artificial intelligence, particularly when players operate under incomplete information. While communication helps, it is not always feasible. In this paper, we explore how effective coordination can be achieved without verbal communication, relying solely on observing each other's actions. Our method enables an agent to develop a strategy by interpreting its partner's action sequences as intent signals, constructing a finite-state transducer built from deterministic finite automata, one for each possible action the agent can take. Experiments show that these strategies significantly outperform uncoordinated ones and closely match the performance of coordinating via direct communication.