In multi-agent systems, agents possess only local observations of the environment. Communication between teammates becomes crucial for enhancing coordination. Past research has primarily focused on encoding local information into embedding messages which are unintelligible to humans. We find that using these messages in agent's policy learning leads to brittle policies when tested on out-of-distribution initial states. We present an approach to multi-agent coordination, where each agent is equipped with the capability to integrate its (history of) observations, actions and messages received into a Common Operating Picture (COP) and disseminate the COP. This process takes into account the dynamic nature of the environment and the shared mission. We conducted experiments in the StarCraft2 environment to validate our approach. Our results demonstrate the efficacy of COP integration, and show that COP-based training leads to robust policies compared to state-of-the-art Multi-Agent Reinforcement Learning (MARL) methods when faced with out-of-distribution initial states.

Distributed optimization has experienced a significant surge in interest due to its wide-ranging applications in distributed learning and adaptation. While various scenarios, such as shared-memory, local-memory, and consensus-based approaches, have been extensively studied in isolation, there remains a need for further exploration of their interconnections. This paper specifically concentrates on a scenario where agents collaborate toward a unified mission while potentially having distinct tasks. Each agent's actions can potentially impact other agents through interactions. Within this context, the objective for the agents is to optimize their local parameters based on the aggregate of local reward functions, where only local zeroth-order oracles are available. Notably, the learning process is asynchronous, meaning that agents update and query their zeroth-order oracles asynchronously while communicating with other agents subject to bounded but possibly random communication delays. This paper presents theoretical convergence analyses and establishes a convergence rate for the proposed approach. Furthermore, it addresses the relevant issue of deep learning-based resource allocation in communication networks and conducts numerical experiments in which agents, acting as transmitters, collaboratively train their individual (possibly unique) policies to maximize a common performance metric.

Policy-Space Response Oracles (PSRO) is an influential algorithm framework for approximating a Nash Equilibrium (NE) in multi-agent non-transitive games. Many previous studies have been trying to promote policy diversity in PSRO. A major weakness in existing diversity metrics is that a more diverse (according to their diversity metrics) population does not necessarily mean (as we proved in the paper) a better approximation to a NE. To alleviate this problem, we propose a new diversity metric, the improvement of which guarantees a better approximation to a NE. Meanwhile, we develop a practical and well-justified method to optimize our diversity metric using only state-action samples. By incorporating our diversity regularization into the best response solving in PSRO, we obtain a new PSRO variant, Policy Space Diversity PSRO (PSD-PSRO). We present the convergence property of PSD-PSRO. Empirically, extensive experiments on various games demonstrate that PSD-PSRO is more effective in producing significantly less exploitable policies than state-of-the-art PSRO variants.

We consider a generic decentralized constrained optimization problem over static, directed communication networks, where each agent has exclusive access to only one convex, differentiable, local objective term and one convex constraint set. For this setup, we propose a novel decentralized algorithm, called DAGP (Double Averaging and Gradient Projection), based on local gradients, projection onto local constraints, and local averaging. We achieve global optimality through a novel distributed tracking technique we call distributed null projection. Further, we show that DAGP can be used to solve unconstrained problems with non-differentiable objective terms with a problem reduction scheme. Assuming only smoothness of the objective terms, we study the convergence of DAGP and establish sub-linear rates of convergence in terms of feasibility, consensus, and optimality, with no extra assumption (e.g. strong convexity). For the analysis, we forego the difficulties of selecting Lyapunov functions by proposing a new methodology of convergence analysis in optimization problems, which we refer to as aggregate lower-bounding. To demonstrate the generality of this method, we also provide an alternative convergence proof for the standard gradient descent algorithm with smooth functions. Finally, we present numerical results demonstrating the effectiveness of our proposed method in both constrained and unconstrained problems. In particular, we propose a distributed scheme by DAGP for the optimal transport problem with superior performance and speed.

We propose a novel approach to the statistical analysis of stochastic simulation models and, especially, agent-based models (ABMs). Our main goal is to provide fully automated, model-independent and tool-supported techniques and algorithms to inspect simulations and perform counterfactual analysis. Our approach: (i) is easy-to-use by the modeller, (ii) improves reproducibility of results, (iii) optimizes running time given the modeller's machine, (iv) automatically chooses the number of required simulations and simulation steps to reach user-specified statistical confidence, and (v) automates a variety of statistical tests. In particular, our techniques are designed to distinguish the transient dynamics of the model from its steady-state behaviour (if any), estimate properties in both 'phases', and provide indications on the (non-)ergodic nature of the simulated processes - which, in turn, allows one to gauge the reliability of a steady-state analysis. Estimates are equipped with statistical guarantees, allowing for robust comparisons across computational experiments. To demonstrate the effectiveness of our approach, we apply it to two models from the literature: a large-scale macro-financial ABM and a small scale prediction market model. Compared to prior analyses of these models, we obtain new insights and we are able to identify and fix some erroneous conclusions.

Embodied AI models often employ off the shelf vision backbones like CLIP to encode their visual observations. Although such general purpose representations encode rich syntactic and semantic information about the scene, much of this information is often irrelevant to the specific task at hand. This introduces noise within the learning process and distracts the agent's focus from task-relevant visual cues. Inspired by selective attention in humans-the process through which people filter their perception based on their experiences, knowledge, and the task at hand-we introduce a parameter-efficient approach to filter visual stimuli for embodied AI. Our approach induces a task-conditioned bottleneck using a small learnable codebook module. This codebook is trained jointly to optimize task reward and acts as a task-conditioned selective filter over the visual observation. Our experiments showcase state-of-the-art performance for object goal navigation and object displacement across 5 benchmarks, ProcTHOR, ArchitecTHOR, RoboTHOR, AI2-iTHOR, and ManipulaTHOR. The filtered representations produced by the codebook are also able generalize better and converge faster when adapted to other simulation environments such as Habitat. Our qualitative analyses show that agents explore their environments more effectively and their representations retain task-relevant information like target object recognition while ignoring superfluous information about other objects. Code and pretrained models are available at our project website: https://embodied-codebook.github.io.

Theory of Mind (ToM) is a fundamental cognitive architecture that endows humans with the ability to attribute mental states to others. Humans infer the desires, beliefs, and intentions of others by observing their behavior and, in turn, adjust their actions to facilitate better interpersonal communication and team collaboration. In this paper, we investigated trust-aware robot policy with the theory of mind in a multiagent setting where a human collaborates with a robot against another human opponent. We show that by only focusing on team performance, the robot may resort to the reverse psychology trick, which poses a significant threat to trust maintenance. The human's trust in the robot will collapse when they discover deceptive behavior by the robot. To mitigate this problem, we adopt the robot theory of mind model to infer the human's trust beliefs, including true belief and false belief (an essential element of ToM). We designed a dynamic trust-aware reward function based on different trust beliefs to guide the robot policy learning, which aims to balance between avoiding human trust collapse due to robot reverse psychology. The experimental results demonstrate the importance of the ToM-based robot policy for human-robot trust and the effectiveness of our robot ToM-based robot policy in multiagent interaction settings.

Social intelligence manifests the capability, often referred to as the Theory of Mind (ToM), to discern others' behavioral intentions, beliefs, and other mental states. ToM is especially important in multi-agent and human-machine interaction environments because each agent needs to understand the mental states of other agents in order to better respond, interact, and collaborate. Recent research indicates that the ToM model possesses the capability to infer beliefs, intentions, and anticipate future observations and actions; nonetheless, its deployment in tackling intricate tasks remains notably limited. The challenges arise when the number of agents increases, the environment becomes more complex, and interacting with the environment and predicting the mental state of each other becomes difficult and time consuming. To overcome such limits, we take inspiration from the Theory of Collective Mind (ToCM) mechanism, predicting observations of all other agents into a unified but plural representation and discerning how our own actions affect this mental state representation. Based on this foundation, we construct an imaginative space to simulate the multi-agent interaction process, thus improving the efficiency of cooperation among multiple agents in complex decision-making environments. In various cooperative tasks with different numbers of agents, the experimental results highlight the superior cooperative efficiency and performance of our approach compared to the Multi-Agent Reinforcement Learning (MARL) baselines. We achieve consistent boost on SNN- and DNN-based decision networks, and demonstrate that ToCM's inferences about others' mental states can be transferred to new tasks for quickly and flexible adaptation.

This paper considers optimal traffic signal control in smart cities, which has been taken as a complex networked system control problem. Given the interacting dynamics among traffic lights and road networks, attaining controller adaptivity and scalability stands out as a primary challenge. Capturing the spatial-temporal correlation among traffic lights under the framework of Multi-Agent Reinforcement Learning (MARL) is a promising solution. Nevertheless, existing MARL algorithms ignore effective information aggregation which is fundamental for improving the learning capacity of decentralized agents. In this paper, we design a new decentralized control architecture with improved environmental observability to capture the spatial-temporal correlation. Specifically, we first develop a topology-aware information aggregation strategy to extract correlation-related information from unstructured data gathered in the road network. Particularly, we transfer the road network topology into a graph shift operator by forming a diffusion process on the topology, which subsequently facilitates the construction of graph signals. A diffusion convolution module is developed, forming a new MARL algorithm, which endows agents with the capabilities of graph learning. Extensive experiments based on both synthetic and real-world datasets verify that our proposal outperforms existing decentralized algorithms.

Neural MMO 2.0 is a massively multi-agent environment for reinforcement learning research. The key feature of this new version is a flexible task system that allows users to define a broad range of objectives and reward signals. We challenge researchers to train agents capable of generalizing to tasks, maps, and opponents never seen during training. Neural MMO features procedurally generated maps with 128 agents in the standard setting and support for up to. Version 2.0 is a complete rewrite of its predecessor with three-fold improved performance and compatibility with CleanRL. We release the platform as free and open-source software with comprehensive documentation available at neuralmmo.github.io and an active community Discord. To spark initial research on this new platform, we are concurrently running a competition at NeurIPS 2023.