Interactive Decision Making for Autonomous V ehicles in Dense Traffic David Isele 1 Abstract -- Dense urban traffic environments can produce situations where accurate prediction and dynamic models are insufficient for successful autonomous vehicle motion planning. We investigate how an autonomous agent can safely negotiate with other traffic participants, enabling the agent to handle potential deadlocks. Specifically we consider merges where the gap between cars is smaller than the size of the ego vehicle. We propose a game theoretic framework capable of generating and responding to interactive behaviors. Our main contribution is to show how game-tree decision making can be executed by an autonomous vehicle, including approximations and reasoning that make the tree-search computationally tractable. Additionally, to test our model we develop a stochastic rule-based traffic agent capable of generating interactive behaviors that can be used as a benchmark for simulating traffic participants in a crowded merge setting. I NTRODUCTION Much of the long tail around autonomous driving behavior relates to complex interactions between self-interested agents. Since other traffic participants exhibit a great deal of variety and are often neither purely adversarial, nor purely cooperative, it can be difficult to reason about their behavior. However this type of reasoning is essential to numerous traffic situations in congested traffic such as overcrowded merge scenarios depicted in Figure 1.

This paper investigates a population-based training regime based on game-theoretic principles called Policy-Spaced Response Oracles (PSRO). PSRO is general in the sense that it (1) encompasses well-known algorithms such as fictitious play and double oracle as special cases, and (2) in principle applies to general-sum, many-player games. Despite this, prior studies of PSRO have been focused on two-player zero-sum games, a regime wherein Nash equilibria are tractably computable. In moving from two-player zero-sum games to more general settings, computation of Nash equilibria quickly becomes infeasible. Here, we extend the theoretical underpinnings of PSRO by considering an alternative solution concept, {\alpha}-Rank, which is unique (thus faces no equilibrium selection issues, unlike Nash) and tractable to compute in general-sum, many-player settings. We establish convergence guarantees in several games classes, and identify links between Nash equilibria and {\alpha}-Rank. We demonstrate the competitive performance of {\alpha}-Rank-based PSRO against an exact Nash solver-based PSRO in 2-player Kuhn and Leduc Poker. We then go beyond the reach of prior PSRO applications by considering 3- to 5-player poker games, yielding instances where {\alpha}-Rank achieves faster convergence than approximate Nash solvers, thus establishing it as a favorable general games solver. We also carry out an initial empirical validation in MuJoCo soccer, illustrating the feasibility of the proposed approach in another complex domain.

Most previous studies on multi-agent reinforcement learning focus on deriving decentralized and cooperative policies to maximize a common reward and rarely consider the transferability of trained policies to new tasks. This prevents such policies from being applied to more complex multi-agent tasks. To resolve these limitations, we propose a model that conducts both representation learning for multiple agents using hierarchical graph attention network and policy learning using multi-agent actor-critic. The hierarchical graph attention network is specially designed to model the hierarchical relationships among multiple agents that either cooperate or compete with each other to derive more advanced strategic policies. Two attention networks, the inter-agent and inter-group attention layers, are used to effectively model individual and group level interactions, respectively. The two attention networks have been proven to facilitate the transfer of learned policies to new tasks with different agent compositions and allow one to interpret the learned strategies. Empirically, we demonstrate that the proposed model outperforms existing methods in several mixed cooperative and competitive tasks.

Machine learning has enabled significant benefits in diverse fields, but, with a few exceptions, has had limited impact on computer architecture. Recent work, however, has explored broader applicability for design, optimization, and simulation. Notably, machine learning based strategies often surpass prior state-of-the-art analytical, heuristic, and human-expert approaches. This paper reviews machine learning applied system-wide to simulation and run-time optimization, and in many individual components, including memory systems, branch predictors, networks-on-chip, and GPUs. The paper further analyzes current practice to highlight useful design strategies and identify areas for future work, based on optimized implementation strategies, opportune extensions to existing work, and ambitious long term possibilities. Taken together, these strategies and techniques present a promising future for increasingly automated architectural design.

Under the headline "AI safety", a wide-reaching issue is being discussed, whether in the future some "superhuman artificial intelligence" / "superintelligence" could could pose a threat to humanity. In addition, the late Steven Hawking warned that the rise of robots may be disastrous for mankind. A major concern is that even benevolent superhuman artificial intelligence (AI) may become seriously harmful if its given goals are not exactly aligned with ours, or if we cannot specify precisely its objective function. Metaphorically, this is compared to king Midas in Greek mythology, who expressed the wish that everything he touched should turn to gold, but obviously this wish was not specified precisely enough. In our view, this sounds like requirements problems and the challenge of their precise formulation. (To our best knowledge, this has not been pointed out yet.) As usual in requirements engineering (RE), ambiguity or incompleteness may cause problems. In addition, the overall issue calls for a major RE endeavor, figuring out the wishes and the needs with regard to a superintelligence, which will in our opinion most likely be a very complex software-intensive system based on AI. This may even entail theoretically defining an extended requirements problem.

Cooperative learning in heterogeneous teams refers to learning methods in which teams are organised both to accomplish academic tasks and for individuals to gain knowledge. Competencies, personality and the gender of team members are key factors that influence team performance. Here, we introduce a team composition problem, the so-called synergistic team composition problem (STCP), which incorporates such key factors when arranging teams. Thus, the goal of the STCP is to partition a set of individuals into a set of synergistic teams: teams that are diverse in personality and gender and whose members cover all required competencies to complete a task. Furthermore, the STCP requires that all teams are balanced in that they are expected to exhibit similar performances when completing the task. We propose two efficient algorithms to solve the STCP . Our first algorithm is based on a linear programming formulation and is appropriate to solve small instances of the problem. Our second algorithm is an anytime heuristic that is effective for large instances of the STCP . Finally, we thoroughly study the computational properties of both algorithms in an educational context when grouping students in a classroom into teams using actual-world data. Keywords: team composition, exact algorithms, heuristic algorithms, optimisation, coalition formation 1. Introduction Active learning refers to a broad range of teaching techniques that engage students to participate in all learning activities in the classes. Typically, active learning strategies involve a substantial amount of students working together within teams. They do not only acquire and retain the information better but also are more content with their classes [2]. Nevertheless, not all teams facilitate learning. For team-based learning to be effective, every team composed in the classroom needs to be heterogeneous, i.e. diverse in individuals' characteristics. Furthermore, having some significantly weaker teams and some significantly stronger teams is undesirable. Hence, the distribution of teams in a classroom must be balanced in the sense that all teams are more or less equally strong. Even though much research in the industrial, organisational, and educational psychology fields investigated what are the predictors of team success, to the best of our knowledge, there are no computational models to build teams for a given task that are broadly used in the classrooms. Frequently studied individual characteristics that influence team performance are competencies, personality traits, and gender [3, 4, 5, 6]. Some of those characteristics were also acknowledged by multiagent systems (MAS) research. The most studied characteristic in MAS research are competencies [7, 8, 9, 10, 11].

We consider the problem of controlling a large fleet of drones to deliver packages simultaneously across broad urban areas. To conserve their limited flight range, drones can seamlessly hop between and ride on top of public transit vehicles (e.g., buses and trams). We design a novel comprehensive algorithmic framework that strives to minimize the maximum time to complete any delivery. We address the multifaceted complexity of the problem through a two-layer approach. First, the upper layer assigns drones to package delivery sequences with a provably near-optimal polynomial-time task allocation algorithm. Then, the lower layer executes the allocation by periodically routing the fleet over the transit network while employing efficient bounded-suboptimal multi-agent pathfinding techniques tailored to our setting. We present extensive experiments supporting the efficiency of our approach on settings with up to $200$ drones, $5000$ packages, and large transit networks of up to $8000$ stops in San Francisco and the Washington DC area. Our results show that the framework can compute solutions within a few seconds (up to $2$ minutes for the largest settings) on commodity hardware, and that drones travel up to $450 \%$ of their flight range by using public transit.

Given a certain complexity level, humanized agents may select from a wide range of possible tasks, with each activity corresponding to a transient goal. In general there will be no overarching credit assignment scheme allowing to compare available options with respect to expected utilities. For this situation we propose a task selection framework that is based on time allocation via emotional stationarity (TAES). Emotions are argued to correspond to abstract criteria, such as satisfaction, challenge and boredom, along which activities that have been carried out can be evaluated. The resulting timeline of experienced emotions is then compared with the `character' of the agent, which is defined in terms of a preferred distribution of emotional states. The long-term goal of the agent, to align experience with character, is achieved by optimizing the frequency for selecting the individual tasks. Upon optimization, the statistics of emotion experience becomes stationary.

We investigate the combination of actor-critic reinforcement learning algorithms with uniform large-scale experience replay and propose solutions for two challenges: (a) efficient actor-critic learning with experience replay (b) stability of very off-policy learning. We employ those insights to accelerate hyper-parameter sweeps in which all participating agents run concurrently and share their experience via a common replay module. To this end we analyze the bias-variance tradeoffs in V-trace, a form of importance sampling for actor-critic methods. Based on our analysis, we then argue for mixing experience sampled from replay with on-policy experience, and propose a new trust region scheme that scales effectively to data distributions where V-trace becomes unstable. We provide extensive empirical validation of the proposed solution. We further show the benefits of this setup by demonstrating state-of-the-art data efficiency on Atari among agents trained up until 200M environment frames.

To coordinate with other systems, agents must be able to determine what the systems are currently doing and predict what they will be doing in the future---plan and goal recognition. There are many methods for plan and goal recognition, but they assume a passive observer that continually monitors the target system. Real-world domains, where information gathering has a cost (e.g., moving a camera or a robot, or time taken away from another task), will often require a more active observer. We propose to combine goal recognition with other observer tasks in order to obtain \emph{active goal recognition} (AGR). We discuss this problem and provide a model and preliminary experimental results for one form of this composite problem. As expected, the results show that optimal behavior in AGR problems balance information gathering with other actions (e.g., task completion) such as to achieve all tasks jointly and efficiently. We hope that our formulation opens the door for extensive further research on this interesting and realistic problem.