We address the problem of building agents whose goal is to satisfy out-of distribution (OOD) multi-task instructions expressed in temporal logic (TL) by using deep reinforcement learning (DRL). Recent works provided evidence that the deep learning architecture is a key feature when teaching a DRL agent to solve OOD tasks in TL. Yet, the studies on their performance are still limited. In this work, we analyse various state-of-the-art (SOTA) architectures that include generalisation mechanisms such as relational layers, the soft-attention mechanism, or hierarchical configurations, when generalising safety-aware tasks expressed in TL. Most importantly, we present a novel deep learning architecture that induces agents to generate latent representations of their current goal given both the human instruction and the current observation from the environment. We find that applying our proposed configuration to SOTA architectures yields significantly stronger performance when executing new tasks in OOD environments.

Value alignment in AI has emerged as one of the basic principles that should govern autonomous AI systems. It essentially states that a system's goals and behaviour should be aligned with human values. But how to ensure value alignment? In this paper we first provide a formal model to represent values through preferences and ways to compute value aggregations; i.e. preferences with respect to a group of agents and/or preferences with respect to sets of values. Value alignment is then defined, and computed, for a given norm with respect to a given value through the increase/decrease that it results in the preferences of future states of the world. We focus on norms as it is norms that govern behaviour, and as such, the alignment of a given system with a given value will be dictated by the norms the system follows.

DNA-based nanonetworks have a wide range of promising use cases, especially in the field of medicine. With a large set of agents, a partially observable stochastic environment, and noisy observations, such nanoscale systems can be modelled as a decentralised, partially observable, Markov decision process (DecPOMDP). As the agent set is a dominating factor, this paper presents (i) lifted DecPOMDPs, partitioning the agent set into sets of indistinguishable agents, reducing the worst-case space required, and (ii) a nanoscale medical system as an application. Future work turns to solving and implementing lifted DecPOMDPs.

But when groups are involved, with many people grabbing the wheel at once, we often find ourselves in a fruitless stalemate headed for disaster, or worse, lurching off the road and into a ditch, seemingly just to spite ourselves. It turns out that Mother Nature has been working on this problem for hundreds of millions of years, evolving countless species that make effective decisions in large groups. A human business team trying to select the ideal location for a new factory would face a similarly complex problem and find it very difficult to choose optimally, and yet simple honeybees achieve this. They do so by forming real-time systems that efficiently combine the diverse perspectives of the hundreds of scout bees that explored the available options, enabling group deliberation that considers their differing levels of conviction until they converge on a single unified decision. It enables groups of all sizes to connect over the internet and deliberate as a unified system, pushing and pulling on decisions while swarming algorithms monitor their actions and reactions.

The current landscape of multi-agent expert imitation is broadly dominated by two families of algorithms - Behavioral Cloning (BC) and Adversarial Imitation Learning (AIL). BC approaches suffer from compounding errors, as they ignore the sequential decision-making nature of the trajectory generation problem. Furthermore, they cannot effectively model multi-modal behaviors. While AIL methods solve the issue of compounding errors and multi-modal policy training, they are plagued with instability in their training dynamics. In this work, we address this issue by introducing a novel self-supervised loss that encourages the discriminator to approximate a richer reward function. We employ our method to train a graph-based multi-agent actor-critic architecture that learns a centralized policy, conditioned on a learned latent interaction graph. We show that our method (SS-MAIL) outperforms prior state-of-the-art methods on real-world prediction tasks, as well as on custom-designed synthetic experiments. We prove that SS-MAIL is part of the family of AIL methods by providing a theoretical connection to cost-regularized apprenticeship learning. Moreover, we leverage the self-supervised formulation to introduce a novel teacher forcing-based curriculum (Trajectory Forcing) that improves sample efficiency by progressively increasing the length of the generated trajectory. The SS-MAIL framework improves multi-agent imitation capabilities by stabilizing the policy training, improving the reward shaping capabilities, as well as providing the ability for modeling multi-modal trajectories.

Federated Learning (FL) over wireless multi-hop edge computing networks, i.e., multi-hop FL, is a cost-effective distributed on-device deep learning paradigm. This paper presents FedEdge simulator, a high-fidelity Linux-based simulator, which enables fast prototyping, sim-to-real code, and knowledge transfer for multi-hop FL systems. FedEdge simulator is built on top of the hardware-oriented FedEdge experimental framework with a new extension of the realistic physical layer emulator. This emulator exploits trace-based channel modeling and dynamic link scheduling to minimize the reality gap between the simulator and the physical testbed. Our initial experiments demonstrate the high fidelity of the FedEdge simulator and its superior performance on sim-to-real knowledge transfer in reinforcement learning-optimized multi-hop FL.

We address the problem of routing a team of drones and trucks over large-scale urban road networks. To conserve their limited flight energy, drones can use trucks as temporary modes of transit en route to their own destinations. Such coordination can yield significant savings in total vehicle distance traveled, i.e., truck travel distance and drone flight distance, compared to operating drones and trucks independently. But it comes at the potentially prohibitive computational cost of deciding which trucks and drones should coordinate and when and where it is most beneficial to do so. We tackle this fundamental trade-off by decoupling our overall intractable problem into tractable sub-problems that we solve stage-wise. The first stage solves only for trucks, by computing paths that make them more likely to be useful transit options for drones. The second stage solves only for drones, by routing them over a composite of the road network and the transit network defined by truck paths from the first stage. We design a comprehensive algorithmic framework that frames each stage as a multi-agent path-finding problem and implement two distinct methods for solving them. We evaluate our approach on extensive simulations with up to $100$ agents on the real-world Manhattan road network containing nearly $4500$ vertices and $10000$ edges. Our framework saves on more than $50\%$ of vehicle distance traveled compared to independently solving for trucks and drones, and computes solutions for all settings within $5$ minutes on commodity hardware.

Motion prediction is important for intelligent driving systems, providing the future distributions of road agent behaviors and supporting various decision making tasks. Existing motion predictors are often optimized and evaluated via task-agnostic measures based on prediction accuracy. Such measures fail to account for the use of prediction in downstream tasks, and could result in sub-optimal task performance. We propose a task-informed motion prediction framework that jointly reasons about prediction accuracy and task utility, to better support downstream tasks through its predictions. The task utility function does not require the full task information, but rather a specification of the utility of the task, resulting in predictors that serve a wide range of downstream tasks. We demonstrate our framework on two use cases of task utilities, in the context of autonomous driving and parallel autonomy, and show the advantage of task-informed predictors over task-agnostic ones on the Waymo Open Motion dataset.

We aim to develop an AI agent that can watch video clips and have a conversation with human about the video story. Developing video understanding intelligence is a significantly challenging task, and evaluation methods for adequately measuring and analyzing the progress of AI agent are lacking as well. In this paper, we propose the Video Turing Test to provide effective and practical assessments of video understanding intelligence as well as human-likeness evaluation of AI agents. We define a general format and procedure of the Video Turing Test and present a case study to confirm the effectiveness and usefulness of the proposed test.

Let's face it, we humans make a lot of bad decisions. And even when we are deeply aware that our decisions are hurting ourselves -- like destroying our environment or propagating inequality -- we seem collectively helpless to correct course. It is exasperating, like watching a car heading for a brick wall with a driver that seems unwilling or unable to turn the wheel. Ironically, as individuals, we are not nearly as dysfunctional, most of us turning the wheel as needed to navigate our daily lives. But when groups are involved, with many people grabbing the wheel at once, we often find ourselves in a fruitless stalemate headed for disaster, or worse, lurching off the road and into a ditch, seemingly just to spite ourselves.