For robots to seamlessly interact with humans, we first need to make sure that humans and robots understand one another. Diverse algorithms have been developed to enable robots to learn from humans (i.e., transferring information from humans to robots). In parallel, visual, haptic, and auditory communication interfaces have been designed to convey the robot's internal state to the human (i.e., transferring information from robots to humans). Prior research often separates these two directions of information transfer, and focuses primarily on either learning algorithms or communication interfaces. By contrast, in this review we take an interdisciplinary approach to identify common themes and emerging trends that close the loop between learning and communication. Specifically, we survey state-of-the-art methods and outcomes for communicating a robot's learning back to the human teacher during human-robot interaction. This discussion connects human-in-the-loop learning methods and explainable robot learning with multi-modal feedback systems and measures of human-robot interaction. We find that -- when learning and communication are developed together -- the resulting closed-loop system can lead to improved human teaching, increased human trust, and human-robot co-adaptation. The paper includes a perspective on several of the interdisciplinary research themes and open questions that could advance how future robots communicate their learning to everyday operators. Finally, we implement a selection of the reviewed methods in a case study where participants kinesthetically teach a robot arm. This case study documents and tests an integrated approach for learning in ways that can be communicated, conveying this learning across multi-modal interfaces, and measuring the resulting changes in human and robot behavior. See videos of our case study here: https://youtu.be/EXfQctqFzWs

In this study, we explore the application of Large Language Models (LLMs) in "Jubensha" (Chinese murder mystery role-playing games), a novel area in AI-driven gaming. We introduce the first Chinese dataset specifically for Jubensha, including character scripts and game rules, to foster AI agent development in this complex narrative environment. Our work also presents a unique multi-agent interaction framework using LLMs, allowing AI agents to autonomously engage in the game, enhancing the dynamics of Jubensha gameplay. To evaluate these AI agents, we developed specialized methods targeting their mastery of case information and reasoning skills. Furthermore, we incorporated the latest advancements in in-context learning to improve the agents' performance in critical aspects like information gathering, murderer detection, and logical reasoning. The experimental results validate the effectiveness of our proposed methods. This work aims to offer a fresh perspective on understanding LLM capabilities and establish a new benchmark for evaluating large language model-based agents to researchers in the field.

Group decision making is an important field with interesting applications in various disciplines, among which environmental economics. Group decision, often requires that all or the majority of agents in the group agree to a single proposal or opinion, i.e. consensus. This is particularly true in cases where there is no coercion involved in the implementation of the decision made, so that the implementation of the decision depends on the good will, or rather the acceptance of the common decision by all members of the group. To make the discussion more concrete we consider the following generic situation: Assume that a group of agents, G, has to reach a common decision concerning policies regarding a future contingency X. Policies may refer for instance to the cost of abatement measures for protection against X, which clearly require the acceptance of a commonly acceptable estimate for the value of X by every member of the group as well as the acceptance of a commonly acceptably discount factor. Typically, different member of the group will have different valuations for X, therefore report different costs for the adverse effects of X. Moreover, different members of the group will have different discount rates for calculating the present value of the future adverse effect X.

Route planning for a fleet of vehicles is an important task in applications such as package delivery, surveillance, or transportation. This problem is usually modeled as a Combinatorial Optimization problem named as Team Orienteering Problem. The most popular Team Orienteering Problem solvers are mainly based on either linear programming, which provides accurate solutions by employing a large computation time that grows with the size of the problem, or heuristic methods, which usually find suboptimal solutions in a shorter amount of time. In this paper, a multi-agent route planning system capable of solving the Team Orienteering Problem in a very fast and accurate manner is presented. The proposed system is based on a centralized Transformer neural network that can learn to encode the scenario (modeled as a graph) and the context of the agents to provide fast and accurate solutions. Several experiments have been performed to demonstrate that the presented system can outperform most of the state-of-the-art works in terms of computation speed. In addition, the code is publicly available at http://gti.ssr.upm.es/data.

For effective collaboration between humans and intelligent agents that employ machine learning for decision-making, humans must understand what agents can and cannot do to avoid over/under-reliance. A solution to this problem is adjusting human reliance through communication using reliance calibration cues (RCCs) to help humans assess agents' capabilities. Previous studies typically attempted to calibrate reliance by continuously presenting RCCs, and when an agent should provide RCCs remains an open question. To answer this, we propose Pred-RC, a method for selectively providing RCCs. Pred-RC uses a cognitive reliance model to predict whether a human will assign a task to an agent. By comparing the prediction results for both cases with and without an RCC, Pred-RC evaluates the influence of the RCC on human reliance. We tested Pred-RC in a human-AI collaboration task and found that it can successfully calibrate human reliance with a reduced number of RCCs.

The goal of this research is to devise guaranteed defense policies that allow to protect a given region from the entrance of smart mobile invaders by detecting them using a team of defending agents equipped with identical line sensors. By designing cooperative defense strategies that ensure all invaders are detected, conditions on the defenders' speed are derived. Successful accomplishment of the defense task implies invaders with a known limit on their speed cannot slip past the defenders and enter the guarded region undetected. The desired outcome of the defense protocols is to defend the area and additionally to expand it as much as possible. Expansion becomes possible if the defenders' speed exceeds a critical speed that is necessary to only defend the initial region. We present results on the total search time, critical speeds and maximal expansion possible for two types of novel pincer-movement defense processes, circular and spiral, for any even number of defenders. The proposed spiral process allows to detect invaders at nearly the lowest theoretically optimal speed, and if this speed is exceeded, it also allows to expand the protected region almost to the maximal area.

Interpretable policy learning seeks to estimate intelligible decision policies from observed actions; however, existing models fall short by forcing a tradeoff between accuracy and interpretability. This tradeoff limits data-driven interpretations of human decision-making process. e.g. to audit medical decisions for biases and suboptimal practices, we require models of decision processes which provide concise descriptions of complex behaviors. Fundamentally, existing approaches are burdened by this tradeoff because they represent the underlying decision process as a universal policy, when in fact human decisions are dynamic and can change drastically with contextual information. Thus, we propose Contextualized Policy Recovery (CPR), which re-frames the problem of modeling complex decision processes as a multi-task learning problem in which complex decision policies are comprised of context-specific policies. CPR models each context-specific policy as a linear observation-to-action mapping, and generates new decision models $\textit{on-demand}$ as contexts are updated with new observations. CPR is compatible with fully offline and partially observable decision environments, and can be tailored to incorporate any recurrent black-box model or interpretable decision model. We assess CPR through studies on simulated and real data, achieving state-of-the-art performance on the canonical tasks of predicting antibiotic prescription in intensive care units ($+22\%$ AUROC vs. previous SOTA) and predicting MRI prescription for Alzheimer's patients ($+7.7\%$ AUROC vs. previous SOTA). With this improvement in predictive performance, CPR closes the accuracy gap between interpretable and black-box methods for policy learning, allowing high-resolution exploration and analysis of context-specific decision models.

ML inference workflows often require low latency and high throughput, yet we lack good options for addressing this need. Techniques that reduce latency in other streaming settings (such as caching and optimization-driven scheduling) are of limited value because ML data dependencies are often very large and can change dramatically depending on the triggering event. In this work, we propose a novel correlation grouping mechanism that makes it easier for developers to express application-specific data access correlations, enabling coordinated management of data objects in server clusters hosting streaming inference tasks. Experiments based on a latency-sensitive ML-based application confirm the limitations of standard techniques while showing that our solution yields dramatically better performance. The proposed mechanism is able to maintain significantly lower and more consistent latency, achieves higher node utilization as workload and scale-out increase, and yet requires only minor changes to the code implementing the application.

Several different fairness notions have been introduced in the context of fair allocation of goods. In this manuscript, we compare between some fairness notions that are used in settings in which agents have arbitrary (perhaps unequal) entitlements to the goods. This includes the proportional share, the anyprice share, the weighted maximin share, weighted envy freeness, maximum weight Nash social welfare and competitive equilibrium. We perform this comparison in two settings, that of a divisible homogeneous good and arbitrary valuations, and that of indivisible goods and additive valuations. Different fairness notions are not always compatible with each other, and might dictate selecting different allocations. The purpose of our work is to clarify various properties of fairness notions, so as to allow, when needed, to make an educated choice among them. Also, such a study may motivate introducing new fairness notions, or modifications to existing fairness notions. Among other properties, we introduce definitions for monotonicity that postulate that having higher entitlement should be better to the agent than having lower entitlement. Some monotonicity notions, such as population monotonicity and weight monotonicity, appeared in previous work, but we prefer to consider other monotonicity properties that we refer to as global monotonicity and individual monotonicity. We find that some of the fairness notions (but not all) violate our monotonicity properties in a strong sense, that we refer to as the inversion paradox. Under this paradox, a fairness notion enforces that the value received by an agent decreases when the entitlement of the agent increases.

The challenging task of Vision-and-Language Navigation (VLN) requires embodied agents to follow natural language instructions to reach a goal location or object (e.g. `walk down the hallway and turn left at the piano'). For agents to complete this task successfully, they must be able to ground objects referenced into the instruction (e.g.`piano') into the visual scene as well as ground directional phrases (e.g.`turn left') into actions. In this work we ask the following question -- to what degree are spatial and directional language cues informing the navigation model's decisions? We propose a series of simple masking experiments to inspect the model's reliance on different parts of the instruction. Surprisingly we uncover that certain top performing models rely only on the noun tokens of the instructions. We propose two training methods to alleviate this concerning limitation.