This paper deals with the following common type of scenario: several users engage in a repeated online auction, where each of them is assisted by a learning agent. A typical example is advertisers that compete for ad-slots: typically, each of these advertisers fills in his key parameters into some advertiser-facing website, and then this website's "agent" participates on the advertiser's behalf in a sequence of auctions for ad slots. Often, the auction's platform designer provides this agent as its advertiser-facing user interface. In cases where the platform's agent does not optimize sufficiently well for the advertiser (but rather, say, for the auctioneer) one would expect some other company to provide a better (for the advertiser) agent. A typical learning algorithm of this type will have at its core some regret-minimization algorithm [33, 9], such as multiplicative weights (see [2] and references therein) or some variant of fictitious play [12, 54], such as follow the perturbed leader [31, 40]. In particular, for ad auctions, there is empirical data showing that bids are largely consistent with no-regret learning [46, 52].

It has long been believed that Chess is the \emph{Drosophila} of Artificial Intelligence (AI). Studying Chess can productively provide valid knowledge about complex systems. Although remarkable progress has been made on solving Chess, the geometrical landscape of Chess in the strategy space is still mysterious. Judging on AI-generated strategies, researchers hypothesised that the strategy space of Chess possesses a spinning top geometry, with the upright axis representing the \emph{transitive} dimension (e.g., A beats B, B beats C, A beats C), and the radial axis representing the \emph{non-transitive} dimension (e.g., A beats B, B beats C, C beats A). However, it is unclear whether such a hypothesis holds for real-world strategies. In this paper, we quantify the non-transitivity in Chess through real-world data from human players. Specifically, we performed two ways of non-transitivity quantifications -- Nash Clustering and counting the number of Rock-Paper-Scissor cycles -- on over one billion match data from Lichess and FICS. Our findings positively indicate that the strategy space occupied by real-world Chess strategies demonstrates a spinning top geometry, and more importantly, there exists a strong connection between the degree of non-transitivity and the progression of a Chess player's rating. In particular, high degrees of non-transitivity tend to prevent human players from making progress on their Elo rating, whereas progressions are easier to make at the level of ratings where the degree of non-transitivity is lower. Additionally, we also investigate the implication of the degree of non-transitivity for population-based training methods. By considering \emph{fixed-memory Fictitious Play} as a proxy, we reach the conclusion that maintaining large-size and diverse populations of strategies is imperative to training effective AI agents in solving Chess types of games.

Humans find AI to be a frustrating teammate when playing a cooperative game together, posing challenges for "teaming intelligence," study shows. When it comes to games such as chess or Go, artificial intelligence (AI) programs have far surpassed the best players in the world. These "superhuman" AIs are unmatched competitors, but perhaps harder than competing against humans is collaborating with them. Can the same technology get along with people? In a new study, MIT Lincoln Laboratory researchers sought to find out how well humans could play the cooperative card game Hanabi with an advanced AI model trained to excel at playing with teammates it has never met before.

This paper presents a hierarchical framework to solve the multi-robot temporal task planning problem. We assume that each robot has its individual task specification and the robots have to jointly satisfy a global collaborative task specification, both described in linear temporal logic. Specifically, a central server firstly extracts and decomposes a collaborative task sequence from the automaton corresponding to the collaborative task specification, and allocates the subtasks in the sequence to robots. The robots can then synthesize their initial execution strategies based on locally constructed product automatons, combining the assigned collaborative tasks and their individual task specifications. Furthermore, we propose a distributed execution strategy adjusting mechanism to iteratively improve the time efficiency, by reducing wait time in collaborations caused by potential synchronization constraints. We prove the completeness of the proposed framework under assumptions, and analyze its time complexity and optimality. Extensive simulation results verify the scalability and optimization efficiency of the proposed method.

As more and more AI agents are used in practice, it is time to think about how to make these agents fully autonomous so that they can learn by themselves in a self-motivated and self-supervised manner rather than being retrained periodically on the initiation of human engineers using expanded training data. As the real-world is an open environment with unknowns or novelties, detecting novelties or unknowns, gathering ground-truth training data, and incrementally learning the unknowns make the agent more and more knowledgeable and powerful over time. The key challenge is how to automate the process so that it is carried out on the agent's own initiative and through its own interactions with humans and the environment. Since an AI agent usually has a performance task, characterizing each novelty becomes necessary so that the agent can formulate an appropriate response to adapt its behavior to cope with the novelty and to learn from it to improve its future responses and task performance. This paper proposes a theoretic framework for this learning paradigm to promote the research of building self-initiated open world learning agents.

Microservice-based architectures enable different aspects of web applications to be created and updated independently, even after deployment. Associated technologies such as service mesh provide application-level fault resilience through attribute configurations that govern the behavior of request-response service -- and the interactions among them -- in the presence of failures. While this provides tremendous flexibility, the configured values of these attributes -- and the relationships among them -- can significantly affect the performance and fault resilience of the overall application. Furthermore, it is impossible to determine the best and worst combinations of attribute values with respect to fault resiliency via testing, due to the complexities of the underlying distributed system and the many possible attribute value combinations. In this paper, we present a model-based reinforcement learning workflow towards service mesh fault resiliency. Our approach enables the prediction of the most significant fault resilience behaviors at a web application-level, scratching from single service to aggregated multi-service management with efficient agent collaborations.

Abstract--Artificial agents are promising for realtime power system operations, particularly, to compute remedial actions for congestion management. Currently, these agents are limited to only autonomously run by themselves. However, autonomous agents will not be deployed any time soon. Operators will still be in charge of taking action in the future. Aiming at designing an assistant for operators, we here consider humans in the loop and propose an original formulation for this problem. We first advance an agent with the ability to send to the operator alarms ahead of time when the proposed actions are of low confidence. We further model the operator's available attention as a budget that decreases when alarms are sent. We present the design and results of our competition "Learning to run a power network with trust" in which we benchmark the ability of submitted agents to send relevant alarms while operating the network to their best.

Autonomous systems are highly complex and present unique challenges for the application of formal methods. Autonomous systems act without human intervention, and are often embedded in a robotic system, so that they can interact with the real world. As such, they exhibit the properties of safety-critical, cyber-physical, hybrid, and real-time systems. This EPTCS volume contains the proceedings for the third workshop on Formal Methods for Autonomous Systems (FMAS 2021), which was held virtually on the 21st and 22nd of October 2021. Like the previous workshop, FMAS 2021 was an online, stand-alone event, as an adaptation to the ongoing COVID-19 restrictions. Despite the challenges this brought, we were determined to build on the success of the previous two FMAS workshops. The goal of FMAS is to bring together leading researchers who are tackling the unique challenges of autonomous systems using formal methods, to present recent and ongoing work. We are interested in the use of formal methods to specify, model, or verify autonomous and/or robotic systems; in whole or in part. We are also interested in successful industrial applications and potential future directions for this emerging application of formal methods.

Undesired bias afflicts both human and algorithmic decision making, and may be especially prevalent when information processing trade-offs incentivize the use of heuristics. One primary example is \textit{statistical discrimination} -- selecting social partners based not on their underlying attributes, but on readily perceptible characteristics that covary with their suitability for the task at hand. We present a theoretical model to examine how information processing influences statistical discrimination and test its predictions using multi-agent reinforcement learning with various agent architectures in a partner choice-based social dilemma. As predicted, statistical discrimination emerges in agent policies as a function of both the bias in the training population and of agent architecture. All agents showed substantial statistical discrimination, defaulting to using the readily available correlates instead of the outcome relevant features. We show that less discrimination emerges with agents that use recurrent neural networks, and when their training environment has less bias. However, all agent algorithms we tried still exhibited substantial bias after learning in biased training populations.

In tabular multi-agent reinforcement learning with average-cost criterion, a team of agents sequentially interacts with the environment and observes local incentives. We focus on the case that the global reward is a sum of local rewards, the joint policy factorizes into agents' marginals, and full state observability. To date, few global optimality guarantees exist even for this simple setting, as most results yield convergence to stationarity for parameterized policies in large/possibly continuous spaces. To solidify the foundations of MARL, we build upon linear programming (LP) reformulations, for which stochastic primal-dual methods yields a model-free approach to achieve \emph{optimal sample complexity} in the centralized case. We develop multi-agent extensions, whereby agents solve their local saddle point problems and then perform local weighted averaging. We establish that the sample complexity to obtain near-globally optimal solutions matches tight dependencies on the cardinality of the state and action spaces, and exhibits classical scalings with respect to the network in accordance with multi-agent optimization. Experiments corroborate these results in practice.