Safe reinforcement learning (RL) agents accomplish given tasks while adhering to specific constraints. Employing constraints expressed via easily-understandable human language offers considerable potential for real-world applications due to its accessibility and non-reliance on domain expertise. Previous safe RL methods with natural language constraints typically adopt a recurrent neural network, which leads to limited capabilities when dealing with various forms of human language input. Furthermore, these methods often require a ground-truth cost function, necessitating domain expertise for the conversion of language constraints into a well-defined cost function that determines constraint violation. To address these issues, we proposes to use pre-trained language models (LM) to facilitate RL agents' comprehension of natural language constraints and allow them to infer costs for safe policy learning. Through the use of pre-trained LMs and the elimination of the need for a ground-truth cost, our method enhances safe policy learning under a diverse set of human-derived free-form natural language constraints. Experiments on grid-world navigation and robot control show that the proposed method can achieve strong performance while adhering to given constraints. The usage of pre-trained LMs allows our method to comprehend complicated constraints and learn safe policies without the need for ground-truth cost at any stage of training or evaluation. Extensive ablation studies are conducted to demonstrate the efficacy of each part of our method.

Multi-Agent Policy Gradient (MAPG) has made significant progress in recent years. However, centralized critics in state-of-the-art MAPG methods still face the centralized-decentralized mismatch (CDM) issue, which means sub-optimal actions by some agents will affect other agent's policy learning. While using individual critics for policy updates can avoid this issue, they severely limit cooperation among agents. To address this issue, we propose an agent topology framework, which decides whether other agents should be considered in policy gradient and achieves compromise between facilitating cooperation and alleviating the CDM issue. The agent topology allows agents to use coalition utility as learning objective instead of global utility by centralized critics or local utility by individual critics. To constitute the agent topology, various models are studied. We propose Topology-based multi-Agent Policy gradiEnt (TAPE) for both stochastic and deterministic MAPG methods. We prove the policy improvement theorem for stochastic TAPE and give a theoretical explanation for the improved cooperation among agents. Experiment results on several benchmarks show the agent topology is able to facilitate agent cooperation and alleviate CDM issue respectively to improve performance of TAPE. Finally, multiple ablation studies and a heuristic graph search algorithm are devised to show the efficacy of the agent topology.

In emergency search and rescue scenarios, the quick location of trapped people is essential. However, disasters can render the Global Positioning System (GPS) unusable. Unmanned aerial vehicles (UAVs) with localization devices can serve as mobile anchors due to their agility and high line-of-sight (LoS) probability. Nonetheless, the number of available UAVs during the initial stages of disaster relief is limited, and innovative methods are needed to quickly plan UAV trajectories to locate non-uniformly distributed dynamic targets while ensuring localization accuracy. To address this challenge, we design a single UAV localization method without hovering, use the maximum likelihood estimation (MLE) method to estimate the location of mobile users and define the upper bound of the localization error by considering users' movement.Combining this localization method and localization error-index, we utilize the enhanced particle swarm optimization (EPSO) algorithm and edge access strategy to develop a low complexity localization-oriented adaptive trajectory planning algorithm. Simulation results demonstrate that our method outperforms other baseline algorithms, enabling faster localization without compromising localization accuracy.

The paper begins by exploring the rationality of ethical trust as a foundational concept. This involves distinguishing between trust and trustworthiness and delving into scenarios where trust is both rational and moral. It lays the groundwork for understanding the complexities of trust dynamics in decision-making scenarios. Following this theoretical groundwork, we introduce an agent-based simulation framework that investigates these dynamics of ethical trust, specifically in the context of a disaster response scenario. These agents, utilizing emotional models like Plutchik's Wheel of Emotions and memory learning mechanisms, are tasked with allocating limited resources in disaster-affected areas. The model, which embodies the principles discussed in the first section, integrates cognitive load management, Big Five personality traits, and structured interactions within networked or hierarchical settings. It also includes feedback loops and simulates external events to evaluate their impact on the formation and evolution of trust among agents. Through our simulations, we demonstrate the intricate interplay of cognitive, emotional, and social factors in ethical decision-making. These insights shed light on the behaviors and resilience of trust networks in crisis situations, emphasizing the role of rational and moral considerations in the development of trust among autonomous agents. This study contributes to the field by offering an understanding of trust dynamics in socio-technical systems and by providing a robust, adaptable framework capable of addressing ethical dilemmas in disaster response and beyond. The implementation of the algorithms presented in this paper is available at this GitHub repository: \url{https://github.com/abbas-tari/ethical-trust-cognitive-modeling}.

Hedonic Games (HGs) are a classical framework modeling coalition formation of strategic agents guided by their individual preferences. According to these preferences, it is desirable that a coalition structure (i.e. a partition of agents into coalitions) satisfies some form of stability. The most well-known and natural of such notions is arguably core-stability. Informally, a partition is core-stable if no subset of agents would like to deviate by regrouping in a so-called core-blocking coalition. Unfortunately, core-stable partitions seldom exist and even when they do, it is often computationally intractable to find one. To circumvent these problems, we propose the notion of $\varepsilon$-fractional core-stability, where at most an $\varepsilon$-fraction of all possible coalitions is allowed to core-block. It turns out that such a relaxation may guarantee both existence and polynomial-time computation. Specifically, we design efficient algorithms returning an $\varepsilon$-fractional core-stable partition, with $\varepsilon$ exponentially decreasing in the number of agents, for two fundamental classes of HGs: Simple Fractional and Anonymous. From a probabilistic point of view, being the definition of $\varepsilon$-fractional core equivalent to requiring that uniformly sampled coalitions core-block with probability lower than $\varepsilon$, we further extend the definition to handle more complex sampling distributions. Along this line, when valuations have to be learned from samples in a PAC-learning fashion, we give positive and negative results on which distributions allow the efficient computation of outcomes that are $\varepsilon$-fractional core-stable with arbitrarily high confidence.

Conversational agents have been gaining increasing popularity in recent years. Influenced by the widespread adoption of task-oriented agents such as Apple Siri and Amazon Alexa, these agents are being deployed into various applications to enhance user experience. Although these agents promote "ask me anything" functionality, they are typically built to focus on a single or finite set of expertise. Given that complex tasks often require more than one expertise, this results in the users needing to learn and adopt multiple agents. One approach to alleviate this is to abstract the orchestration of agents in the background. However, this removes the option of choice and flexibility, potentially harming the ability to complete tasks. In this paper, we explore these different interaction experiences (one agent for all) vs (user choice of agents) for conversational AI. We design prototypes for each, systematically evaluating their ability to facilitate task completion. Through a series of conducted user studies, we show that users have a significant preference for abstracting agent orchestration in both system usability and system performance. Additionally, we demonstrate that this mode of interaction is able to provide quality responses that are rated within 1% of human-selected answers.

We have used agent-based modeling as our numerical method to artificially simulate a dynamic real economy where agents are rational maximizers of an objective function of Cobb-Douglas type. The economy is characterised by heterogeneous agents, acting out of local or imperfect information, monopolistic competition, perfect product differentiation, allowance for increasing returns to scale technology and trade in disequilibrium. An algorithm for economic activity in each period is devised and a general purpose open source agent-based model is developed which allows for counterfactual inquiries, testing out treatments, analysing causality of various economic processes, outcomes and studying emergent properties. 10,000 simulations, with 10 firms and 80 consumers are run with varying parameters and the results show that from only a few initial conditions the economy reaches equilibrium while in most of the other cases it remains in perpetual disequilibrium. It also shows that from a few initial conditions the economy reaches a disaster where all the consumer wealth falls to zero or only a single producer remains. Furthermore, from some initial conditions, an ideal economy with high wage rate, high consumer utility and no unemployment is also reached. It was also observed that starting from an equal endowment of wealth in consumers and in producers, inequality emerged in the economy. In majority of the cases most of the firms(6-7) shut down because they were not profitable enough and only a few firms remained. Our results highlight that all these varying outcomes are possible for a decentralized market economy with rational optimizing agents.

Recent advances in Multi-Agent Reinforcement Learning have prompted the modeling of intricate interactions between agents in simulated environments. In particular, the predator-prey dynamics have captured substantial interest and various simulations been tailored to unique requirements. To prevent further time-intensive developments, we introduce Aquarium, a comprehensive Multi-Agent Reinforcement Learning environment for predator-prey interaction, enabling the study of emergent behavior. Aquarium is open source and offers a seamless integration of the PettingZoo framework, allowing a quick start with proven algorithm implementations. It features physics-based agent movement on a two-dimensional, edge-wrapping plane. The agent-environment interaction (observations, actions, rewards) and the environment settings (agent speed, prey reproduction, predator starvation, and others) are fully customizable. Besides a resource-efficient visualization, Aquarium supports to record video files, providing a visual comprehension of agent behavior. To demonstrate the environment's capabilities, we conduct preliminary studies which use PPO to train multiple prey agents to evade a predator. In accordance to the literature, we find Individual Learning to result in worse performance than Parameter Sharing, which significantly improves coordination and sample-efficiency.

Multi-Agent Reinforcement Learning is becoming increasingly more important in times of autonomous driving and other smart industrial applications. Simultaneously a promising new approach to Reinforcement Learning arises using the inherent properties of quantum mechanics, reducing the trainable parameters of a model significantly. However, gradient-based Multi-Agent Quantum Reinforcement Learning methods often have to struggle with barren plateaus, holding them back from matching the performance of classical approaches. We build upon an existing approach for gradient free Quantum Reinforcement Learning and propose three genetic variations with Variational Quantum Circuits for Multi-Agent Reinforcement Learning using evolutionary optimization. We evaluate our genetic variations in the Coin Game environment and also compare them to classical approaches. We showed that our Variational Quantum Circuit approaches perform significantly better compared to a neural network with a similar amount of trainable parameters. Compared to the larger neural network, our approaches archive similar results using $97.88\%$ less parameters.

Train timetable rescheduling (TTR) aims to promptly restore the original operation of trains after unexpected disturbances or disruptions. Currently, this work is still done manually by train dispatchers, which is challenging to maintain performance under various problem instances. To mitigate this issue, this study proposes a reinforcement learning-based approach to TTR, which makes the following contributions compared to existing work. First, we design a simple directed graph to represent the TTR problem, enabling the automatic extraction of informative states through graph neural networks. Second, we reformulate the construction process of TTR's solution, not only decoupling the decision model from the problem size but also ensuring the generated scheme's feasibility. Third, we design a learning curriculum for our model to handle the scenarios with different levels of delay. Finally, a simple local search method is proposed to assist the learned decision model, which can significantly improve solution quality with little additional computation cost, further enhancing the practical value of our method. Extensive experimental results demonstrate the effectiveness of our method. The learned decision model can achieve better performance for various problems with varying degrees of train delay and different scales when compared to handcrafted rules and state-of-the-art solvers.