Meta-training agents with memory has been shown to culminate in Bayes-optimal agents, which casts Bayes-optimality as the implicit solution to a numerical optimization problem rather than an explicit modeling assumption. Bayes-optimal agents are risk-neutral, since they solely attune to the expected return, and ambiguity-neutral, since they act in new situations as if the uncertainty were known. This is in contrast to risk-sensitive agents, which additionally exploit the higher-order moments of the return, and ambiguity-sensitive agents, which act differently when recognizing situations in which they lack knowledge. Humans are also known to be averse to ambiguity and sensitive to risk in ways that aren't Bayes-optimal, indicating that such sensitivity can confer advantages, especially in safety-critical situations. How can we extend the meta-learning protocol to generate risk- and ambiguity-sensitive agents? The goal of this work is to fill this gap in the literature by showing that risk- and ambiguity-sensitivity also emerge as the result of an optimization problem using modified meta-training algorithms, which manipulate the experience-generation process of the learner. We empirically test our proposed meta-training algorithms on agents exposed to foundational classes of decision-making experiments and demonstrate that they become sensitive to risk and ambiguity.

Collision avoidance is a crucial task in vision-guided autonomous navigation. Solutions based on deep reinforcement learning (DRL) has become increasingly popular. In this work, we proposed several novel agent state and reward function designs to tackle two critical issues in DRL-based navigation solutions: 1) smoothness of the trained flight trajectories; and 2) model generalization to handle unseen environments. Formulated under a DRL framework, our model relies on margin reward and smoothness constraints to ensure UAVs fly smoothly while greatly reducing the chance of collision. The proposed smoothness reward minimizes a combination of first-order and second-order derivatives of flight trajectories, which can also drive the points to be evenly distributed, leading to stable flight speed. To enhance the agent's capability of handling new unseen environments, two practical setups are proposed to improve the invariance of both the state and reward function when deploying in different scenes. Experiments demonstrate the effectiveness of our overall design and individual components.

Rendering to their abilities to provide ubiquitous connectivity, flexibly and cost effectively, unmanned aerial vehicles (UAVs) have been getting more and more research attention. To take the UAVs' performance to the next level, however, they need to be merged with some other technologies like non-orthogonal multiple access (NOMA) and millimeter wave (mmWave), which both promise high spectral efficiency (SE). As managing UAVs efficiently may not be possible using model-based techniques, another key innovative technology that UAVs will inevitably need to leverage is artificial intelligence (AI). Designing an AI-based technique that adaptively allocates radio resources and places UAVs in 3D space to meet certain communication objectives, however, is a tough row to hoe. In this paper, we propose a neuroevolution of augmenting topologies NEAT framework, referred to as ProSky, to manage NOMA-mmWave-UAV networks. ProSky exhibits a remarkable performance improvement over a model-based method. Moreover, ProSky learns 5.3 times faster than and outperforms, in both SE and energy efficiency EE while being reasonably fair, a deep reinforcement learning DRL based scheme. The ProSky source code is accessible to use here: https://github.com/Fouzibenfaid/ProSky

Abstract: Autonomous driving functions (ADFs) in public traffic have to comply with complex system requirements that are based on knowledge of experts from different disciplines, e.g., lawyers, safety experts, psychologists. In this paper, we present a research preview regarding the validation of ADFs with respect to such requirements. We investigate the suitability of Traffic Sequence Charts (TSCs) for the formalization of such requirements and present a concept for monitoring system compliance during validation runs. We find TSCs, with their intuitive visual syntax over symbols from the traffic domain, to be a promising choice for the collaborative formalization of such requirements. For an example TSC, we describe the construction of a runtime monitor according to our novel concept that exploits the separation of spatial and temporal aspects in TSCs, and successfully apply the monitor on exemplary runs.

In safety-critical applications of reinforcement learning such as healthcare and robotics, it is often desirable to optimize risk-sensitive objectives that account for tail outcomes rather than expected reward. We prove the first regret bounds for reinforcement learning under a general class of risk-sensitive objectives including the popular CVaR objective. Our theory is based on a novel characterization of the CVaR objective as well as a novel optimistic MDP construction.

Episodic count has been widely used to design a simple yet effective intrinsic motivation for reinforcement learning with a sparse reward. However, the use of episodic count in a high-dimensional state space as well as over a long episode time requires a thorough state compression and fast hashing, which hinders rigorous exploitation of it in such hard and complex exploration environments. Moreover, the interference from task-irrelevant observations in the episodic count may cause its intrinsic motivation to overlook task-related important changes of states, and the novelty in an episodic manner can lead to repeatedly revisit the familiar states across episodes. In order to resolve these issues, in this paper, we propose a learnable hash-based episodic count, which we name LECO, that efficiently performs as a task-specific intrinsic reward in hard exploration problems. In particular, the proposed intrinsic reward consists of the episodic novelty and the task-specific modulation where the former employs a vector quantized variational autoencoder to automatically obtain the discrete state codes for fast counting while the latter regulates the episodic novelty by learning a modulator to optimize the task-specific extrinsic reward. The proposed LECO specifically enables the automatic transition from exploration to exploitation during reinforcement learning. We experimentally show that in contrast to the previous exploration methods LECO successfully solves hard exploration problems and also scales to large state spaces through the most difficult tasks in MiniGrid and DMLab environments.

The use of interactive advice in reinforcement learning scenarios allows for speeding up the learning process for autonomous agents. Current interactive reinforcement learning research has been limited to real-time interactions that offer relevant user advice to the current state only. Moreover, the information provided by each interaction is not retained and instead discarded by the agent after a single use. In this paper, we present a method for retaining and reusing provided knowledge, allowing trainers to give general advice relevant to more than just the current state. Results obtained show that the use of broad-persistent advice substantially improves the performance of the agent while reducing the number of interactions required for the trainer.

No-press Diplomacy is a complex strategy game involving both cooperation and competition that has served as a benchmark for multi-agent AI research. While self-play reinforcement learning has resulted in numerous successes in purely adversarial games like chess, Go, and poker, self-play alone is insufficient for achieving optimal performance in domains involving cooperation with humans. We address this shortcoming by first introducing a planning algorithm we call DiL-piKL that regularizes a reward-maximizing policy toward a human imitationlearned policy. We prove that this is a no-regret learning algorithm under a modified utility function. We then show that DiL-piKL can be extended into a self-play reinforcement learning algorithm we call RL-DiL-piKL that provides a model of human play while simultaneously training an agent that responds well to this human model. We used RL-DiL-piKL to train an agent we name Diplodocus. In a 200-game no-press Diplomacy tournament involving 62 human participants spanning skill levels from beginner to expert, two Diplodocus agents both achieved a higher average score than all other participants who played more than two games, and ranked first and third according to an Elo ratings model. In two-player zero-sum (2p0s) settings, principled self-play algorithms converge to a minimax equilibrium, which in a balanced game ensures that a player will not lose in expectation regardless of the opponent's strategy (Neumann, 1928). This fact has allowed self-play, even without human data, to achieve remarkable success in 2p0s games like chess (Silver et al., 2018), Go (Silver et al., 2017), poker (Bowling et al., 2015; Brown & Sandholm, 2017), and Dota 2 (Berner et al., 2019). In principle, any finite 2p0s game can be solved via self-play given sufficient compute and memory. However, in games involving cooperation, self-play alone no longer guarantees good performance when playing with humans, even with infinite compute and memory. This is because in complex domains there may be arbitrarily many conventions and expectations for how to cooperate, of which humans may use only a small subset (Lerer & Peysakhovich, 2019). The clearest example of this is language. A self-play agent trained from scratch without human data in a cooperative game involving free-form communication channels would almost certainly not converge to using English as the medium of communication. Obviously, such an agent would perform poorly when paired with a human English speaker. Indeed, prior work has shown that naïve extensions of self-play from scratch without human data perform poorly when playing with humans or human-like agents even in dialogue-free domains that involve cooperation rather than just competition, such as the benchmark games no-press Diplomacy (Bakhtin et al., 2021) and Hanabi (Siu et al., 2021; Cui et al., 2021).

We study whether the learning rate $\alpha$, the discount factor $\gamma$ and the reward signal $r$ have an influence on the overestimation bias of the Q-Learning algorithm. Our preliminary results in environments which are stochastic and that require the use of neural networks as function approximators, show that all three parameters influence overestimation significantly. By carefully tuning $\alpha$ and $\gamma$, and by using an exponential moving average of $r$ in Q-Learning's temporal difference target, we show that the algorithm can learn value estimates that are more accurate than the ones of several other popular model-free methods that have addressed its overestimation bias in the past.

Recent progress in reinforcement learning (RL) has started producing generally capable agents that can solve a distribution of complex environments. These agents are typically tested on fixed, human-authored environments. On the other hand, quality diversity (QD) optimization has been proven to be an effective component of environment generation algorithms, which can generate collections of high-quality environments that are diverse in the resulting agent behaviors. However, these algorithms require potentially expensive simulations of agents on newly generated environments. We propose Deep Surrogate Assisted Generation of Environments (DSAGE), a sample-efficient QD environment generation algorithm that maintains a deep surrogate model for predicting agent behaviors in new environments. Results in two benchmark domains show that DSAGE significantly outperforms existing QD environment generation algorithms in discovering collections of environments that elicit diverse behaviors of a state-of-the-art RL agent and a planning agent. Our source code and videos are available at https://dsagepaper.github.io/.