Modifying Neural Networks in Adversarial Agents of Multi-agent Reinforcement Learning Systems | IEEE Conference Publication | IEEE Xplore