Decision makers often aim to learn a treatment assignment policy under a capacity constraint on the number of agents that they can treat. When agents can respond strategically to such policies, competition arises, complicating estimation of the optimal policy. In this paper, we study capacity-constrained treatment assignment in the presence of such interference. We consider a dynamic model where the decision maker allocates treatments at each time step and heterogeneous agents myopically best respond to the previous treatment assignment policy. When the number of agents is large but finite, we show that the threshold for receiving treatment under a given policy converges to the policy's mean-field equilibrium threshold. Based on this result, we develop a consistent estimator for the policy gradient. In simulations and a semi-synthetic experiment with data from the National Education Longitudinal Study of 1988, we demonstrate that this estimator can be used for learning capacity-constrained policies in the presence of strategic behavior.

Multi-agent reinforcement learning (MARL) research is inherently computationally expensive and it is often difficult to obtain a sufficient number of experiment samples to test hypotheses and make robust statistical claims. Furthermore, MARL algorithms are typically complex in their design and can be tricky to implement correctly. These aspects of MARL present a difficult challenge when it comes to creating useful software for advanced research. Our criteria for such software is that it should be simple enough to use to implement new ideas quickly, while at the same time be scalable and fast enough to test those ideas in a reasonable amount of time. In this preliminary technical report, we introduce Mava, a research library for MARL written purely in JAX, that aims to fulfill these criteria. We discuss the design and core features of Mava, and demonstrate its use and performance across a variety of environments. In particular, we show Mava's substantial speed advantage, with improvements of 10-100x compared to other popular MARL frameworks, while maintaining strong performance. This allows for researchers to test ideas in a few minutes instead of several hours. Finally, Mava forms part of an ecosystem of libraries that seamlessly integrate with each other to help facilitate advanced research in MARL. We hope Mava will benefit the community and help drive scientifically sound and statistically robust research in the field. The open-source repository for Mava is available at https://github.com/instadeepai/Mava.

We consider regret minimization in a general collaborative multi-agent multi-armed bandit model, in which each agent faces a finite set of arms and may communicate with other agents through a central controller. The optimal arm for each agent in this model is the arm with the largest expected mixed reward, where the mixed reward of each arm is a weighted average of its rewards across all agents, making communication among agents crucial. While near-optimal sample complexities for best arm identification are known under this collaborative model, the question of optimal regret remains open. In this work, we address this problem and propose the first algorithm with order optimal regret bounds under this collaborative bandit model. Furthermore, we show that only a small constant number of expected communication rounds is needed.

This paper presents a safety-critical approach to the coordinated control of cooperative robots locomoting in the presence of fixed (holonomic) constraints. To this end, we leverage control barrier functions (CBFs) to ensure the safe cooperation of the robots while maintaining a desired formation and avoiding obstacles. The top-level planner generates a set of feasible trajectories, accounting for both kinematic constraints between the robots and physical constraints of the environment. This planner leverages CBFs to ensure safety-critical coordination control, i.e., guarantee safety of the collaborative robots during locomotion. The middle-level trajectory planner incorporates interconnected single rigid body (SRB) dynamics to generate optimal ground reaction forces (GRFs) to track the safety-ensured trajectories from the top-level planner while addressing the interconnection dynamics between agents. Distributed low-level controllers generate whole-body motion to follow the prescribed optimal GRFs while ensuring the friction cone condition at each end of the stance legs. The effectiveness of the approach is demonstrated through numerical simulations and experimentally on a pair of quadrupedal robots.

Parameter sharing, as an important technique in multi-agent systems, can effectively solve the scalability issue in large-scale agent problems. However, the effectiveness of parameter sharing largely depends on the environment setting. When agents have different identities or tasks, naive parameter sharing makes it difficult to generate sufficiently differentiated strategies for agents. Inspired by research pertaining to the brain in biology, we propose a novel parameter sharing method. It maps each type of agent to different regions within a shared network based on their identity, resulting in distinct subnetworks. Therefore, our method can increase the diversity of strategies among different agents without introducing additional training parameters. Through experiments conducted in multiple environments, our method has shown better performance than other parameter sharing methods.

This project proposes and compares a new way to optimise Super Mario Bros. (SMB) environment where the control is in hand of two approaches, namely, Genetic Algorithm (MarioGA) and NeuroEvolution (MarioNE). Not only we learn playing SMB using these techniques, but also optimise it with constrains of collection of coins and finishing levels. Firstly, we formalise the SMB agent to maximize the total value of collected coins (reward) and maximising the total distance traveled (reward) in order to finish the level faster (time penalty) for both the algorithms. Secondly, we study MarioGA and its evaluation function (fitness criteria) including its representation methods, crossover used, mutation operator formalism, selection method used, MarioGA loop, and few other parameters. Thirdly, MarioNE is applied on SMB where a population of ANNs with random weights is generated, and these networks control Marios actions in the game. Fourth, SMB is further constrained to complete the task within the specified time, rebirths (deaths) within the limit, and performs actions or moves within the maximum allowed moves, while seeking to maximize the total coin value collected. This ensures an efficient way of finishing SMB levels. Finally, we provide a fivefold comparative analysis by plotting fitness plots, ability to finish different levels of world 1, and domain adaptation (transfer learning) of the trained models.

A core function of intelligence is grounding, which is the process of connecting the natural language and abstract knowledge to the internal representation of the real world in an intelligent being, e.g., a human. Human cognition is grounded in our sensorimotor experiences in the external world and subjective feelings in our internal world. We use languages to communicate with each other and the languages are grounded on our shared sensorimotor experiences and feelings. Without this shard grounding, it is impossible for us to understand each other because all natural languages are highly abstract and are only able to describe a tiny portion of what has happened or is happening in the real world. Although grounding at high or abstract levels has been studied in different fields and applications, to our knowledge, limited systematic work at fine-grained levels has been done. With the rapid progress of large language models (LLMs), it is imperative that we have a sound understanding of grounding in order to move to the next level of intelligence. It is also believed that grounding is necessary for Artificial General Intelligence (AGI). This paper makes an attempt to systematically study this problem.

Motion prediction is a crucial task in autonomous driving, and one of its major challenges lands in the multimodality of future behaviors. Many successful works have utilized mixture models which require identification of positive mixture components, and correspondingly fall into two main lines: prediction-based and anchor-based matching. The prediction clustering phenomenon in prediction-based matching makes it difficult to pick representative trajectories for downstream tasks, while the anchor-based matching suffers from a limited regression capability. In this paper, we introduce a novel paradigm, named Evolving and Distinct Anchors (EDA), to define the positive and negative components for multimodal motion prediction based on mixture models. We enable anchors to evolve and redistribute themselves under specific scenes for an enlarged regression capacity. Furthermore, we select distinct anchors before matching them with the ground truth, which results in impressive scoring performance. Our approach enhances all metrics compared to the baseline MTR, particularly with a notable relative reduction of 13.5% in Miss Rate, resulting in state-of-the-art performance on the Waymo Open Motion Dataset. Code is available at https://github.com/Longzhong-Lin/EDA.

This paper explores the potential of a multidisciplinary approach to testing and aligning artificial general intelligence (AGI) and LLMs. Due to the rapid development and wide application of LLMs, challenges such as ethical alignment, controllability, and predictability of these models have become important research topics. This study investigates an innovative simulation-based multi-agent system within a virtual reality framework that replicates the real-world environment. The framework is populated by automated 'digital citizens,' simulating complex social structures and interactions to examine and optimize AGI. Application of various theories from the fields of sociology, social psychology, computer science, physics, biology, and economics demonstrates the possibility of a more human-aligned and socially responsible AGI. The purpose of such a digital environment is to provide a dynamic platform where advanced AI agents can interact and make independent decisions, thereby mimicking realistic scenarios. The actors in this digital city, operated by the LLMs, serve as the primary agents, exhibiting high degrees of autonomy. While this approach shows immense potential, there are notable challenges and limitations, most significantly the unpredictable nature of real-world social dynamics. This research endeavors to contribute to the development and refinement of AGI, emphasizing the integration of social, ethical, and theoretical dimensions for future research.

This paper introduces LLM-MARS, first technology that utilizes a Large Language Model based Artificial Intelligence for Multi-Agent Robot Systems. LLM-MARS enables dynamic dialogues between humans and robots, allowing the latter to generate behavior based on operator commands and provide informative answers to questions about their actions. LLM-MARS is built on a transformer-based Large Language Model, fine-tuned from the Falcon 7B model. We employ a multimodal approach using LoRa adapters for different tasks. The first LoRa adapter was developed by fine-tuning the base model on examples of Behavior Trees and their corresponding commands. The second LoRa adapter was developed by fine-tuning on question-answering examples. Practical trials on a multi-agent system of two robots within the Eurobot 2023 game rules demonstrate promising results. The robots achieve an average task execution accuracy of 79.28% in compound commands. With commands containing up to two tasks accuracy exceeded 90%. Evaluation confirms the system's answers on operators questions exhibit high accuracy, relevance, and informativeness. LLM-MARS and similar multi-agent robotic systems hold significant potential to revolutionize logistics, enabling autonomous exploration missions and advancing Industry 5.0.