Achieving human-AI alignment in complex multi-agent games is crucial for creating trustworthy AI agents that enhance gameplay. We propose a method to evaluate this alignment using an interpretable task-sets framework, focusing on high-level behavioral tasks instead of low-level policies. Our approach has three components. First, we analyze extensive human gameplay data from Xbox's Bleeding Edge (100K+ games), uncovering behavioral patterns in a complex task space. This task space serves as a basis set for a behavior manifold capturing interpretable axes: fight-flight, explore-exploit, and solo-multi-agent. Second, we train an AI agent to play Bleeding Edge using a Generative Pretrained Causal Transformer and measure its behavior. Third, we project human and AI gameplay to the proposed behavior manifold to compare and contrast. This allows us to interpret differences in policy as higher-level behavioral concepts, e.g., we find that while human players exhibit variability in fight-flight and explore-exploit behavior, AI players tend towards uniformity. Furthermore, AI agents predominantly engage in solo play, while humans often engage in cooperative and competitive multi-agent patterns. These stark differences underscore the need for interpretable evaluation, design, and integration of AI in human-aligned applications. Our study advances the alignment discussion in AI and especially generative AI research, offering a measurable framework for interpretable human-agent alignment in multiplayer gaming.

Existing work on trustworthy machine learning (ML) often concentrates on individual aspects of trust, such as fairness or privacy. Additionally, many techniques overlook the distinction between those who train ML models and those responsible for assessing their trustworthiness. To address these issues, we propose a framework that views trustworthy ML as a multi-objective multi-agent optimization problem. This naturally lends itself to a game-theoretic formulation we call regulation games. We illustrate a particular game instance, the SpecGame in which we model the relationship between an ML model builder and fairness and privacy regulators. Regulators wish to design penalties that enforce compliance with their specification, but do not want to discourage builders from participation. Seeking such socially optimal (i.e., efficient for all agents) solutions to the game, we introduce ParetoPlay. This novel equilibrium search algorithm ensures that agents remain on the Pareto frontier of their objectives and avoids the inefficiencies of other equilibria. Simulating SpecGame through ParetoPlay can provide policy guidance for ML Regulation. For instance, we show that for a gender classification application, regulators can enforce a differential privacy budget that is on average 4.0 lower if they take the initiative to specify their desired guarantee first.

This paper considers the contextual multi-armed bandit (CMAB) problem with fairness and privacy guarantees in a federated environment. We consider merit-based exposure as the desired fair outcome, which provides exposure to each action in proportion to the reward associated. We model the algorithm's effectiveness using fairness regret, which captures the difference between fair optimal policy and the policy output by the algorithm. Applying fair CMAB algorithm to each agent individually leads to fairness regret linear in the number of agents. We propose that collaborative -- federated learning can be more effective and provide the algorithm Fed-FairX-LinUCB that also ensures differential privacy. The primary challenge in extending the existing privacy framework is designing the communication protocol for communicating required information across agents. A naive protocol can either lead to weaker privacy guarantees or higher regret. We design a novel communication protocol that allows for (i) Sub-linear theoretical bounds on fairness regret for Fed-FairX-LinUCB and comparable bounds for the private counterpart, Priv-FairX-LinUCB (relative to single-agent learning), (ii) Effective use of privacy budget in Priv-FairX-LinUCB. We demonstrate the efficacy of our proposed algorithm with extensive simulations-based experiments. We show that both Fed-FairX-LinUCB and Priv-FairX-LinUCB achieve near-optimal fairness regret.

There is a sensory gulf between the Earth that humans inhabit and the digital realms in which modern AI agents are created. To develop AI agents that can sense, think, and act as flexibly as humans in real-world settings, it is imperative to bridge the realism gap between the digital and physical worlds. How can we embody agents in an environment as rich and diverse as the one we inhabit, without the constraints imposed by real hardware and control? Towards this end, we introduce V-IRL: a platform that enables agents to scalably interact with the real world in a virtual yet realistic environment. Our platform serves as a playground for developing agents that can accomplish various practical tasks and as a vast testbed for measuring progress in capabilities spanning perception, decision-making, and interaction with real-world data across the entire globe.

We develop a multi-agent reinforcement learning (MARL) algorithm to minimize the total energy consumption of multiple massive MIMO (multiple-input multiple-output) base stations (BSs) in a multi-cell network while preserving the overall quality-of-service (QoS) by making decisions on the multi-level advanced sleep modes (ASMs) and antenna switching of these BSs. The problem is modeled as a decentralized partially observable Markov decision process (DEC-POMDP) to enable collaboration between individual BSs, which is necessary to tackle inter-cell interference. A multi-agent proximal policy optimization (MAPPO) algorithm is designed to learn a collaborative BS control policy. To enhance its scalability, a modified version called MAPPO-neighbor policy is further proposed. Simulation results demonstrate that the trained MAPPO agent achieves better performance compared to baseline policies. Specifically, compared to the auto sleep mode 1 (symbol-level sleeping) algorithm, the MAPPO-neighbor policy reduces power consumption by approximately 8.7% during low-traffic hours and improves energy efficiency by approximately 19% during high-traffic hours, respectively.

Consensus control in multi-agent systems has received significant attention and practical implementation across various domains. However, managing consensus control under unknown dynamics remains a significant challenge for control design due to system uncertainties and environmental disturbances. This paper presents a novel learning-based distributed control law, augmented by an auxiliary dynamics. Gaussian processes are harnessed to compensate for the unknown components of the multi-agent system. For continuous enhancement in predictive performance of Gaussian process model, a data-efficient online learning strategy with a decentralized event-triggered mechanism is proposed. Furthermore, the control performance of the proposed approach is ensured via the Lyapunov theory, based on a probabilistic guarantee for prediction error bounds. To demonstrate the efficacy of the proposed learning-based controller, a comparative analysis is conducted, contrasting it with both conventional distributed control laws and offline learning methodologies.

The combination of self-play and planning has achieved great successes in sequential games, for instance in Chess and Go. However, adapting algorithms such as AlphaZero to simultaneous games poses a new challenge. In these games, missing information about concurrent actions of other agents is a limiting factor as they may select different Nash equilibria or do not play optimally at all. Thus, it is vital to model the behavior of the other agents when interacting with them in simultaneous games. To this end, we propose Albatross: AlphaZero for Learning Bounded-rational Agents and Temperature-based Response Optimization using Simulated Self-play. Albatross learns to play the novel equilibrium concept of a Smooth Best Response Logit Equilibrium (SBRLE), which enables cooperation and competition with agents of any playing strength. We perform an extensive evaluation of Albatross on a set of cooperative and competitive simultaneous perfect-information games. In contrast to AlphaZero, Albatross is able to exploit weak agents in the competitive game of Battlesnake. Additionally, it yields an improvement of 37.6% compared to previous state of the art in the cooperative Overcooked benchmark.

This work presents an innovative learning-based approach to tackle the tracking control problem of Euler-Lagrange multi-agent systems with partially unknown dynamics operating under switching communication topologies. The approach leverages a correlation-aware cooperative algorithm framework built upon Gaussian process regression, which adeptly captures inter-agent correlations for uncertainty predictions. A standout feature is its exceptional efficiency in deriving the aggregation weights achieved by circumventing the computationally intensive posterior variance calculations. Through Lyapunov stability analysis, the distributed control law ensures bounded tracking errors with high probability. Simulation experiments validate the protocol's efficacy in effectively managing complex scenarios, establishing it as a promising solution for robust tracking control in multi-agent systems characterized by uncertain dynamics and dynamic communication structures.

This paper introduces an innovative approach to enhance distributed cooperative learning using Gaussian process (GP) regression in multi-agent systems (MASs). The key contribution of this work is the development of an elective learning algorithm, namely prior-aware elective distributed GP (Pri-GP), which empowers agents with the capability to selectively request predictions from neighboring agents based on their trustworthiness. The proposed Pri-GP effectively improves individual prediction accuracy, especially in cases where the prior knowledge of an agent is incorrect. Moreover, it eliminates the need for computationally intensive variance calculations for determining aggregation weights in distributed GP. Furthermore, we establish a prediction error bound within the Pri-GP framework, ensuring the reliability of predictions, which is regarded as a crucial property in safety-critical MAS applications.

Foundation models have indeed made a profound impact on various fields, emerging as pivotal components that significantly shape the capabilities of intelligent systems. In the context of intelligent vehicles, leveraging the power of foundation models has proven to be transformative, offering notable advancements in visual understanding. Equipped with multi-modal and multi-task learning capabilities, multi-modal multi-task visual understanding foundation models (MM-VUFMs) effectively process and fuse data from diverse modalities and simultaneously handle various driving-related tasks with powerful adaptability, contributing to a more holistic understanding of the surrounding scene. In this survey, we present a systematic analysis of MM-VUFMs specifically designed for road scenes. Our objective is not only to provide a comprehensive overview of common practices, referring to task-specific models, unified multi-modal models, unified multi-task models, and foundation model prompting techniques, but also to highlight their advanced capabilities in diverse learning paradigms. These paradigms include open-world understanding, efficient transfer for road scenes, continual learning, interactive and generative capability. Moreover, we provide insights into key challenges and future trends, such as closed-loop driving systems, interpretability, embodied driving agents, and world models. To facilitate researchers in staying abreast of the latest developments in MM-VUFMs for road scenes, we have established a continuously updated repository at https://github.com/rolsheng/MM-VUFM4DS