Neural networks and games show identical stable states and behaviors

Neural networks and games

Neural and Evolutionary ComputingComputer Science and Game TheoryMachine Learning

Summary

Sometimes, how groups of simple units decide what to do and achieve a common outcome can look very different depending on the system. The authors found that a certain kind of neural network and a certain kind of game where players choose between two options are actually the same in how they settle into stable patterns. These stable patterns in the network match up to stable strategies in the game. This link helps explain how brains store memories and how games find balance points.

What this means in practice

  • For game developers: Design game AI strategies that mimic neural network stable states for predictable player behavior.
  • For cognitive modelers: Use the equivalence to model memory retrieval as equilibrium selection in brain-inspired systems.

A theory result. No direct application yet.

Authors

Philip R. Neary

Abstract

Hopfield neural networks and noncooperative games provide seemingly different approaches to understanding how collective outcomes emerge from local interactions. In a neural network, individual units adjust their states according to local signals; in a game, players choose actions to maximise individual payoffs. I establish an exact equivalence between symmetric binary Hopfield networks and binary-action games that admit a quadratic potential. The equivalence identifies stable network configurations with pure-strategy Nash equilibria, and asynchronous neural updating with best-response dynamics. More broadly, the storage and retrieval of memories in neural networks correspond, respectively, to the design and selection of equilibria in games.