Army researchers have developed a pioneering framework that provides a baseline for the development of collaborative multi-agent systems. The framework is detailed in the survey paper "Survey of recent multi-agent reinforcement learning algorithms utilizing centralized training," which is featured in the SPIE Digital Library. Researchers said the work will support research in reinforcement learning approaches for developing collaborative multi-agent systems such as teams of robots that could work side-by-side with future soldiers. "We propose that the underlying information sharing mechanism plays a critical role in centralized learning for multi-agent systems, but there is limited study of this phenomena within the research community," said Army researcher and computer scientist Dr. Piyush K. Sharma of the U.S. Army Combat Capabilities Development Command, known as DEVCOM, Army Research Laboratory. "We conducted this survey of the state-of-the-art in reinforcement learning algorithms and their information sharing paradigms as a basis for asking fundamental questions on centralized learning for multi-agent systems that would improve their ability to work together."

We study a social learning model in which agents iteratively update their beliefs about the true state of the world using private signals and the beliefs of other agents in a non-Bayesian manner. Some agents are stubborn, meaning they attempt to convince others of an erroneous true state (modeling fake news). We show that while agents learn the true state on short timescales, they "forget" it and believe the erroneous state to be true on longer timescales. Using these results, we devise strategies for seeding stubborn agents so as to disrupt learning, which outperform intuitive heuristics and give novel insights regarding vulnerabilities in social learning.

As AI systems are integrated into high stakes social domains, researchers now examine how to design and operate them in a safe and ethical manner. However, the criteria for identifying and diagnosing safety risks in complex social contexts remain unclear and contested. In this paper, we examine the vagueness in debates about the safety and ethical behavior of AI systems. We show how this vagueness cannot be resolved through mathematical formalism alone, instead requiring deliberation about the politics of development as well as the context of deployment. Drawing from a new sociotechnical lexicon, we redefine vagueness in terms of distinct design challenges at key stages in AI system development. The resulting framework of Hard Choices in Artificial Intelligence (HCAI) empowers developers by 1) identifying points of overlap between design decisions and major sociotechnical challenges; 2) motivating the creation of stakeholder feedback channels so that safety issues can be exhaustively addressed. As such, HCAI contributes to a timely debate about the status of AI development in democratic societies, arguing that deliberation should be the goal of AI Safety, not just the procedure by which it is ensured.

We propose a practical approach to computing market prices and allocations via a deep reinforcement learning policymaker agent, operating in an environment of other learning agents. Compared to the idealized market equilibrium outcome -- which we use as a benchmark -- our policymaker is much more flexible, allowing us to tune the prices with regard to diverse objectives such as sustainability and resource wastefulness, fairness, buyers' and sellers' welfare, etc. To evaluate our approach, we design a realistic market with multiple and diverse buyers and sellers. Additionally, the sellers, which are deep learning agents themselves, compete for resources in a common-pool appropriation environment based on bio-economic models of commercial fisheries. We demonstrate that: (a) The introduced policymaker is able to achieve comparable performance to the market equilibrium, showcasing the potential of such approaches in markets where the equilibrium prices can not be efficiently computed. (b) Our policymaker can notably outperform the equilibrium solution on certain metrics, while at the same time maintaining comparable performance for the remaining ones. (c) As a highlight of our findings, our policymaker is significantly more successful in maintaining resource sustainability, compared to the market outcome, in scarce resource environments.

Measuring and promoting policy diversity is critical for solving games with strong non-transitive dynamics where strategic cycles exist, and there is no consistent winner (e.g., Rock-Paper-Scissors). With that in mind, maintaining a pool of diverse policies via open-ended learning is an attractive solution, which can generate auto-curricula to avoid being exploited. However, in conventional open-ended learning algorithms, there are no widely accepted definitions for diversity, making it hard to construct and evaluate the diverse policies. In this work, we summarize previous concepts of diversity and work towards offering a unified measure of diversity in multi-agent open-ended learning to include all elements in Markov games, based on both Behavioral Diversity (BD) and Response Diversity (RD). At the trajectory distribution level, we re-define BD in the state-action space as the discrepancies of occupancy measures. For the reward dynamics, we propose RD to characterize diversity through the responses of policies when encountering different opponents. We also show that many current diversity measures fall in one of the categories of BD or RD but not both. With this unified diversity measure, we design the corresponding diversity-promoting objective and population effectivity when seeking the best responses in open-ended learning. We validate our methods in both relatively simple games like matrix game, non-transitive mixture model, and the complex \textit{Google Research Football} environment. The population found by our methods reveals the lowest exploitability, highest population effectivity in matrix game and non-transitive mixture model, as well as the largest goal difference when interacting with opponents of various levels in \textit{Google Research Football}.

When in a new situation or geographical location, human drivers have an extraordinary ability to watch others and learn maneuvers that they themselves may have never performed. In contrast, existing techniques for learning to drive preclude such a possibility as they assume direct access to an instrumented ego-vehicle with fully known observations and expert driver actions. However, such measurements cannot be directly accessed for the non-ego vehicles when learning by watching others. Therefore, in an application where data is regarded as a highly valuable asset, current approaches completely discard the vast portion of the training data that can be potentially obtained through indirect observation of surrounding vehicles. Motivated by this key insight, we propose the Learning by Watching (LbW) framework which enables learning a driving policy without requiring full knowledge of neither the state nor expert actions. To increase its data, i.e., with new perspectives and maneuvers, LbW makes use of the demonstrations of other vehicles in a given scene by (1) transforming the ego-vehicle's observations to their points of view, and (2) inferring their expert actions. Our LbW agent learns more robust driving policies while enabling data-efficient learning, including quick adaptation of the policy to rare and novel scenarios. In particular, LbW drives robustly even with a fraction of available driving data required by existing methods, achieving an average success rate of 92% on the original CARLA benchmark with only 30 minutes of total driving data and 82% with only 10 minutes.

In multi-agent reinforcement learning, the inherent non-stationarity of the environment caused by other agents' actions posed significant difficulties for an agent to learn a good policy independently. One way to deal with non-stationarity is agent modeling, by which the agent takes into consideration the influence of other agents' policies. Most existing work relies on predicting other agents' actions or goals, or discriminating between their policies. However, such modeling fails to capture the similarities and differences between policies simultaneously and thus cannot provide useful information when generalizing to unseen policies. To address this, we propose a general method to learn representations of other agents' policies via the joint-action distributions sampled in interactions. The similarities and differences between policies are naturally captured by the policy distance inferred from the joint-action distributions and deliberately reflected in the learned representations. Agents conditioned on the policy representations can well generalize to unseen agents. We empirically demonstrate that our method outperforms existing work in multi-agent tasks when facing unseen agents.

We study fairness through the lens of cooperative multi-agent learning. Our work is motivated by empirical evidence that naive maximization of team reward yields unfair outcomes for individual team members. To address fairness in multi-agent contexts, we introduce team fairness, a group-based fairness measure for multi-agent learning. We then incorporate team fairness into policy optimization -- introducing Fairness through Equivariance (Fair-E), a novel learning strategy that achieves provably fair reward distributions. We then introduce Fairness through Equivariance Regularization (Fair-ER) as a soft-constraint version of Fair-E and show that Fair-ER reaches higher levels of utility than Fair-E and fairer outcomes than policies with no equivariance. Finally, we investigate the fairness-utility trade-off in multi-agent settings.

Algorithms implementing populations of agents which interact with one another and sense their environment may exhibit emergent behavior such as self-organization and swarm intelligence. Here a swarm system, called Databionic swarm (DBS), is introduced which is able to adapt itself to structures of high-dimensional data characterized by distance and/or density-based structures in the data space. By exploiting the interrelations of swarm intelligence, self-organization and emergence, DBS serves as an alternative approach to the optimization of a global objective function in the task of clustering. The swarm omits the usage of a global objective function and is parameter-free because it searches for the Nash equilibrium during its annealing process. To our knowledge, DBS is the first swarm combining these approaches. Its clustering can outperform common clustering methods such as K-means, PAM, single linkage, spectral clustering, model-based clustering, and Ward, if no prior knowledge about the data is available. A central problem in clustering is the correct estimation of the number of clusters. This is addressed by a DBS visualization called topographic map which allows assessing the number of clusters. It is known that all clustering algorithms construct clusters, irrespective of the data set contains clusters or not. In contrast to most other clustering algorithms, the topographic map identifies, that clustering of the data is meaningless if the data contains no (natural) clusters. The performance of DBS is demonstrated on a set of benchmark data, which are constructed to pose difficult clustering problems and in two real-world applications.

Communication is a cornerstone of social interactions, be it with human or artificial intelligence (AI). Yet it can be harmful, depending on the honesty of the exchanged information. To study this, an agent based sociological simulation framework is presented, the reputation game. This illustrates the impact of different communication strategies on the agents' reputation. The game focuses on the trustworthiness of the participating agents, their honesty as perceived by others. In the game, each agent exchanges statements with the others about their own and each other's honesty, which lets their judgments evolve. Various sender and receiver strategies are studied, like sycophant, egocentricity, pathological lying, and aggressiveness for senders as well as awareness and lack thereof for receivers. Minimalist malicious strategies are identified, like being manipulative, dominant, or destructive, which significantly increase reputation at others' costs. Phenomena such as echo chambers, self-deception, deception symbiosis, clique formation, freezing of group opinions emerge from the dynamics. This indicates that the reputation game can be studied for complex group phenomena, to test behavioral hypothesis, and to analyze AI influenced social media. With refined rules it may help to understand social interactions, and to safeguard the design of non-abusive AI systems.