In multi-agent games, the complexity of the environment can grow exponentially as the number of agents increases, so it is particularly challenging to learn good policies when the agent population is large. In this paper, we introduce Evolutionary Population Curriculum (EPC), a curriculum learning paradigm that scales up Multi-Agent Reinforcement Learning (MARL) by progressively increasing the population of training agents in a stage-wise manner. Furthermore, EPC uses an evolutionary approach to fix an objective misalignment issue throughout the curriculum: agents successfully trained in an early stage with a small population are not necessarily the best candidates for adapting to later stages with scaled populations. Concretely, EPC maintains multiple sets of agents in each stage, performs mix-and-match and fine-tuning over these sets and promotes the sets of agents with the best adaptability to the next stage. We implement EPC on a popular MARL algorithm, MADDPG, and empirically show that our approach consistently outperforms baselines by a large margin as the number of agents grows exponentially. The project page is https://sites.google.com/view/epciclr2020.

The next generation of mobile robots needs to be socially-compliant to be accepted by humans. As simple as this task may seem, defining compliance formally is not trivial. Yet, classical reinforcement learning (RL) relies upon hard-coded reward signals. In this work, we go beyond this approach and provide the agent with intrinsic motivation using empowerment. Empowerment maximizes the influence of an agent on its near future and has been shown to be a good model for biological behaviors. It also has been used for artificial agents to learn complicated and generalized actions. Self-empowerment maximizes the influence of an agent on its future. On the contrary, our robot strives for the empowerment of people in its environment, so they are not disturbed by the robot when pursuing their goals. We show that our robot has a positive influence on humans, as it minimizes the travel time and distance of humans while moving efficiently to its own goal. The method can be used in any multi-agent system that requires a robot to solve a particular task involving humans interactions.

Artificial intelligence (AI) is getting a bad reputation. And while Forrester predicts that 1 out of 4 CX professionals will lose their jobs in 2020, it has little to do with employers implementing AI. The real reason is the growing business-criticality of CX. In fact, brands will spend $8 billion more on customer service agents in 2020 than 2019 due to heightened demand and competition for highly skilled agents. The truth is plain and simple: AI isn't replacing contact center agents -- it's helping them step up to be the best they can be and deliver more value than ever.

We consider the problem of using logged data to make predictions about what would happen if we changed the rules of the game' in a multi-agent system. This task is difficult because in many cases we observe actions individuals take but not their private information or their full reward functions. In addition, agents are strategic, so when the rules change, they will also change their actions. They make counterfactual predictions by using observed actions to learn the underlying utility function (a.k.a. This approach imposes heavy assumptions such as the rationality of the agents being observed and a correct model of the environment and agents' utility functions.

Temporal prediction is critical for making intelligent and robust decisions in complex dynamic environments. Motion prediction needs to model the inherently uncertain future which often contains multiple potential outcomes, due to multi-agent interactions and the latent goals of others. Towards these goals, we introduce a probabilistic framework that efficiently learns latent variables to jointly model the multi-step future motions of agents in a scene. Our framework is data-driven and learns semantically meaningful latent variables to represent the multimodal future, without requiring explicit labels. Using a dynamic attention-based state encoder, we learn to encode the past as well as the future interactions among agents, efficiently scaling to any number of agents.

Fairness is essential for human society, contributing to stability and productivity. Similarly, fairness is also the key for many multi-agent systems. Taking fairness into multi-agent learning could help multi-agent systems become both efficient and stable. However, learning efficiency and fairness simultaneously is a complex, multi-objective, joint-policy optimization. To tackle these difficulties, we propose FEN, a novel hierarchical reinforcement learning model.

We show for the first time that it is possible to reconcile in online learning in zero-sum games two seemingly contradictory objectives: vanishing time-average regret and non-vanishing step sizes. This phenomenon, that we coin fast and furious" learning in games, sets a new benchmark about what is possible both in max-min optimization as well as in multi-agent systems. Our analysis does not depend on introducing a carefully tailored dynamic. Instead we focus on the most well studied online dynamic, gradient descent. Similarly, we focus on the simplest textbook class of games, two-agent two-strategy zero-sum games, such as Matching Pennies.

This paper investigates the evaluation of learned multiagent strategies in the incomplete information setting, which plays a critical role in ranking and training of agents. Traditionally, researchers have relied on Elo ratings for this purpose, with recent works also using methods based on Nash equilibria. Unfortunately, Elo is unable to handle intransitive agent interactions, and other techniques are restricted to zero-sum, two-player settings or are limited by the fact that the Nash equilibrium is intractable to compute. Recently, a ranking method called $\alpha$-Rank, relying on a new graph-based game-theoretic solution concept, was shown to tractably apply to general games. However, evaluations based on Elo or $\alpha$-Rank typically assume noise-free game outcomes, despite the data often being collected from noisy simulations, making this assumption unrealistic in practice.

Modelling the dynamics of multi-agent learning has long been an important research topic, but all of the previous works focus on 2-agent settings and mostly use evolutionary game theoretic approaches. In this paper, we study an n-agent setting with n tends to infinity, such that agents learn their policies concurrently over repeated symmetric bimatrix games with some other agents. Using mean field theory, we approximate the effects of other agents on a single agent by an averaged effect. A Fokker-Planck equation that describes the evolution of the probability distribution of Q-values in the agent population is derived. To the best of our knowledge, this is the first time to show the Q-learning dynamics under an n-agent setting can be described by a system of only three equations.

Cooperative multi-agent reinforcement learning often requires decentralised policies, which severely limit the agents' ability to coordinate their behaviour. In this paper, we show that common knowledge between agents allows for complex decentralised coordination. Common knowledge arises naturally in a large number of decentralised cooperative multi-agent tasks, for example, when agents can reconstruct parts of each others' observations. Since agents can independently agree on their common knowledge, they can execute complex coordinated policies that condition on this knowledge in a fully decentralised fashion. We propose multi-agent common knowledge reinforcement learning (MACKRL), a novel stochastic actor-critic algorithm that learns a hierarchical policy tree.