English

Mega-Reward: Achieving Human-Level Play without Extrinsic Rewards

Artificial Intelligence 2019-11-28 v4

Abstract

Intrinsic rewards were introduced to simulate how human intelligence works; they are usually evaluated by intrinsically-motivated play, i.e., playing games without extrinsic rewards but evaluated with extrinsic rewards. However, none of the existing intrinsic reward approaches can achieve human-level performance under this very challenging setting of intrinsically-motivated play. In this work, we propose a novel megalomania-driven intrinsic reward (called mega-reward), which, to our knowledge, is the first approach that achieves human-level performance in intrinsically-motivated play. Intuitively, mega-reward comes from the observation that infants' intelligence develops when they try to gain more control on entities in an environment; therefore, mega-reward aims to maximize the control capabilities of agents on given entities in a given environment. To formalize mega-reward, a relational transition model is proposed to bridge the gaps between direct and latent control. Experimental studies show that mega-reward (i) can greatly outperform all state-of-the-art intrinsic reward approaches, (ii) generally achieves the same level of performance as Ex-PPO and professional human-level scores, and (iii) has also a superior performance when it is incorporated with extrinsic rewards.

Keywords

Cite

@article{arxiv.1905.04640,
  title  = {Mega-Reward: Achieving Human-Level Play without Extrinsic Rewards},
  author = {Yuhang Song and Jianyi Wang and Thomas Lukasiewicz and Zhenghua Xu and Shangtong Zhang and Andrzej Wojcicki and Mai Xu},
  journal= {arXiv preprint arXiv:1905.04640},
  year   = {2019}
}
R2 v1 2026-06-23T09:03:53.723Z