Recently, deep multi-agent reinforcement learning (MARL) has gained significant popularity due to its success in various cooperative multi-agent tasks. However, exploration still remains a challenging problem in MARL due to the partial observability of the agents and the exploration space that can grow exponentially as the number of agents increases. Firstly, in order to address the scalability issue of the exploration space, we define a formation-based equivalence relation on the exploration space and aim to reduce the search space by exploring only meaningful states in different formations. Then, we propose a novel formation-aware exploration (FoX) framework that encourages partially observable agents to visit the states in diverse formations by guiding them to be well aware of their current formation solely based on their own observations. Numerical results show that the proposed FoX framework significantly outperforms the state-of-the-art MARL algorithms on Google Research Football (GRF) and sparse Starcraft II multi-agent challenge (SMAC) tasks.

The interest in using reinforcement learning (RL) controllers in safety-critical applications such as robot navigation around pedestrians motivates the development of additional safety mechanisms. Running RL-enabled systems among uncertain dynamic agents may result in high counts of collisions and failures to reach the goal. The system could be safer if the pre-trained RL policy was uncertainty-informed. For that reason, we propose conformal predictive safety filters that: 1) predict the other agents' trajectories, 2) use statistical techniques to provide uncertainty intervals around these predictions, and 3) learn an additional safety filter that closely follows the RL controller but avoids the uncertainty intervals. We use conformal prediction to learn uncertainty-informed predictive safety filters, which make no assumptions about the agents' distribution. The framework is modular and outperforms the existing controllers in simulation. We demonstrate our approach with multiple experiments in a collision avoidance gym environment and show that our approach minimizes the number of collisions without making overly-conservative predictions.

Identification and analysis of symmetrical patterns in the natural world have led to significant discoveries across various scientific fields, such as the formulation of gravitational laws in physics and advancements in the study of chemical structures. In this paper, we focus on exploiting Euclidean symmetries inherent in certain cooperative multi-agent reinforcement learning (MARL) problems and prevalent in many applications. We begin by formally characterizing a subclass of Markov games with a general notion of symmetries that admits the existence of symmetric optimal values and policies. Motivated by these properties, we design neural network architectures with symmetric constraints embedded as an inductive bias for multi-agent actor-critic methods. This inductive bias results in superior performance in various cooperative MARL benchmarks and impressive generalization capabilities such as zero-shot learning and transfer learning in unseen scenarios with repeated symmetric patterns. The code is available at: https://github.com/dchen48/E3AC.

In interactive multi-agent settings, decision-making complexity arises from agents' interconnected objectives. Dynamic game theory offers a formal framework for analyzing such intricacies. Yet, solving dynamic games and determining Nash equilibria pose computational challenges due to the need of solving coupled optimal control problems. To address this, our key idea is to leverage potential games, which are games with a potential function that allows for the computation of Nash equilibria by optimizing the potential function. We argue that dynamic potential games, can effectively facilitate interactive decision-making in many multi-agent interactions. We will identify structures in realistic multi-agent interactive scenarios that can be transformed into weighted potential dynamic games. We will show that the open-loop Nash equilibria of the resulting weighted potential dynamic game can be obtained by solving a single optimal control problem. We will demonstrate the effectiveness of the proposed method through various simulation studies, showing close proximity to feedback Nash equilibria and significant improvements in solve time compared to state-of-the-art game solvers.

Linear Temporal Logic (LTL) formulas have been used to describe complex tasks for multi-agent systems, with both spatial and temporal constraints. However, since the planning complexity grows exponentially with the number of agents and the length of the task formula, existing applications are mostly limited to small artificial cases. To address this issue, a new planning algorithm is proposed for task formulas specified as sc-LTL formulas. It avoids two common bottlenecks in the model-checking-based planning methods, i.e., (i) the direct translation of the complete task formula to the associated B\"uchi automaton; and (ii) the synchronized product between the B\"uchi automaton and the transition models of all agents. In particular, each conjuncted sub-formula is first converted to the associated R-posets as an abstraction of the temporal dependencies among the subtasks. Then, an efficient algorithm is proposed to compute the product of these R-posets, which retains their dependencies and resolves potential conflicts. Furthermore, the proposed approach is applied to dynamic scenes where new tasks are generated online. It is capable of deriving the first valid plan with a polynomial time and memory complexity w.r.t. the system size and the formula length. Our method can plan for task formulas with a length of more than 60 and a system with more than 35 agents, while most existing methods fail at the formula length of 20. The proposed method is validated on large fleets of service robots in both simulation and hardware experiments.

A key challenge in training generally-capable agents is the design of training tasks that facilitate broad generalization and robustness to environment variations. This challenge motivates the problem setting of Unsupervised Environment Design (UED), whereby a student agent trains on an adaptive distribution of tasks proposed by a teacher agent. A pioneering approach for UED is PAIRED, which uses reinforcement learning (RL) to train a teacher policy to design tasks from scratch, making it possible to directly generate tasks that are adapted to the agent's current capabilities. Despite its strong theoretical backing, PAIRED suffers from a variety of challenges that hinder its practical performance. Thus, state-of-the-art methods currently rely on curation and mutation rather than generation of new tasks. In this work, we investigate several key shortcomings of PAIRED and propose solutions for each shortcoming. As a result, we make it possible for PAIRED to match or exceed state-of-the-art methods, producing robust agents in several established challenging procedurally-generated environments, including a partially-observed maze navigation task and a continuous-control car racing environment. We believe this work motivates a renewed emphasis on UED methods based on learned models that directly generate challenging environments, potentially unlocking more open-ended RL training and, as a result, more general agents.

Whether current or near-term AI systems could be conscious is a topic of scientific interest and increasing public concern. This report argues for, and exemplifies, a rigorous and empirically grounded approach to AI consciousness: assessing existing AI systems in detail, in light of our best-supported neuroscientific theories of consciousness. We survey several prominent scientific theories of consciousness, including recurrent processing theory, global workspace theory, higher-order theories, predictive processing, and attention schema theory. From these theories we derive "indicator properties" of consciousness, elucidated in computational terms that allow us to assess AI systems for these properties. We use these indicator properties to assess several recent AI systems, and we discuss how future systems might implement them. Our analysis suggests that no current AI systems are conscious, but also suggests that there are no obvious technical barriers to building AI systems which satisfy these indicators.

To achieve desired carbon emission reductions, integrating renewable generation and accelerating the adoption of peer-to-peer energy trading is crucial. This is especially important for energy-intensive farming, like dairy farming. However, integrating renewables and peer-to-peer trading presents challenges. To address this, we propose the Multi-Agent Peer-to-Peer Dairy Farm Energy Simulator (MAPDES), enabling dairy farms to participate in peer-to-peer markets. Our strategy reduces electricity costs and peak demand by approximately 30% and 24% respectively, while increasing energy sales by 37% compared to the baseline scenario without P2P trading. This demonstrates the effectiveness of our approach.

Robust coordination skills enable agents to operate cohesively in shared environments, together towards a common goal and, ideally, individually without hindering each other's progress. To this end, this paper presents Coordinated QMIX (CoMIX), a novel training framework for decentralized agents that enables emergent coordination through flexible policies, allowing at the same time independent decision-making at individual level. CoMIX models selfish and collaborative behavior as incremental steps in each agent's decision process. This allows agents to dynamically adapt their behavior to different situations balancing independence and collaboration. Experiments using a variety of simulation environments demonstrate that CoMIX outperforms baselines on collaborative tasks. The results validate our incremental policy approach as effective technique for improving coordination in multi-agent systems.

Efficiently obtaining the up-to-date information in the disaster-stricken area is the key to successful disaster response. Unmanned aerial vehicles (UAVs), workers and cars can collaborate to accomplish sensing tasks, such as data collection, in disaster-stricken areas. In this paper, we explicitly address the route planning for a group of agents, including UAVs, workers, and cars, with the goal of maximizing the task completion rate. We propose MANF-RL-RP, a heterogeneous multi-agent route planning algorithm that incorporates several efficient designs, including global-local dual information processing and a tailored model structure for heterogeneous multi-agent systems. Global-local dual information processing encompasses the extraction and dissemination of spatial features from global information, as well as the partitioning and filtering of local information from individual agents. Regarding the construction of the model structure for heterogeneous multi-agent, we perform the following work. We design the same data structure to represent the states of different agents, prove the Markovian property of the decision-making process of agents to simplify the model structure, and also design a reasonable reward function to train the model. Finally, we conducted detailed experiments based on the rich simulation data. In comparison to the baseline algorithms, namely Greedy-SC-RP and MANF-DNN-RP, MANF-RL-RP has exhibited a significant improvement in terms of task completion rate.