Neural Payoff Machines: Predicting Fair and Stable Payoff Allocations Among Team Members
This work addresses the computational bottleneck in applying cooperative game theory solutions for multi-agent systems, offering a method to accelerate fair reward allocation in domains like team formation and feature attribution, though it is incremental as it builds on existing solution concepts with a learning-based approach.
The paper tackles the computational difficulty of computing fair and stable payoff allocations in cooperative game theory, such as the Shapley value, by distilling these solutions into learned neural network models that generalize to unseen games and larger player counts, with applications in speeding up Shapley value computations for Explainable AI.
In many multi-agent settings, participants can form teams to achieve collective outcomes that may far surpass their individual capabilities. Measuring the relative contributions of agents and allocating them shares of the reward that promote long-lasting cooperation are difficult tasks. Cooperative game theory offers solution concepts identifying distribution schemes, such as the Shapley value, that fairly reflect the contribution of individuals to the performance of the team or the Core, which reduces the incentive of agents to abandon their team. Applications of such methods include identifying influential features and sharing the costs of joint ventures or team formation. Unfortunately, using these solutions requires tackling a computational barrier as they are hard to compute, even in restricted settings. In this work, we show how cooperative game-theoretic solutions can be distilled into a learned model by training neural networks to propose fair and stable payoff allocations. We show that our approach creates models that can generalize to games far from the training distribution and can predict solutions for more players than observed during training. An important application of our framework is Explainable AI: our approach can be used to speed-up Shapley value computations on many instances.