Reinforcement learning has a rich history in neuroscience, from early work on dopamine as a reward prediction error signal for temporal difference learning (Schultz et al., 1997) to recent work suggesting that dopamine could implement a form of 'distributional reinforcement learning' popularized in deep learning (Dabney et al., 2020). Throughout this literature, there has been a tight link between theoretical advances in reinforcement learning and neuroscientific experiments and findings. As a result, the theories describing our experimental data have become increasingly complex and difficult to navigate. In this review, we cover the basic theory underlying classical work in reinforcement learning and build up to an introductory overview of methods used in modern deep reinforcement learning that have found applications in systems neuroscience. We start with an overview of the reinforcement learning problem and classical temporal difference algorithms, followed by a discussion of 'model-free' and 'model-based' reinforcement learning together with methods such as DYNA and successor representations that fall in between these two categories. Throughout these sections, we highlight the close parallels between the machine learning methods and related work in both experimental and theoretical neuroscience. We then provide an introduction to deep reinforcement learning with examples of how these methods have been used to model different learning phenomena in the systems neuroscience literature, such as meta-reinforcement learning (Wang et al., 2018) and distributional reinforcement learning (Dabney et al., 2020). Code that implements the methods discussed in this work and generates the figures is also provided.

C-Procgen provides access to over 200 unique game contexts across 16 games. It allows for detailed configuration of environments, ranging from game mechanics to agent attributes. This makes the procedural generation process, previously a black-box in Procgen, more transparent and adaptable for various research needs. The upgrade enhances dynamic context management and individualized assignments, while maintaining computational efficiency. C-Procgen's controllable contexts make it applicable in diverse reinforcement learning research areas, such as learning dynamics analysis, curriculum learning, and transfer learning. We believe that C-Procgen will fill a gap in the current literature and offer a valuable toolkit for future works.

Tactile information is important for robust performance in robotic tasks that involve physical interaction, such as object manipulation. However, with more data included in the reasoning and control process, modeling behavior becomes increasingly difficult. Deep Reinforcement Learning (DRL) produced promising results for learning complex behavior in various domains, including tactile-based manipulation in robotics. In this work, we present our open-source reinforcement learning environments for the TIAGo service robot. They produce tactile sensor measurements that resemble those of a real sensorised gripper for TIAGo, encouraging research in transfer learning of DRL policies. Lastly, we show preliminary training results of a learned force control policy and compare it to a classical PI controller.

The advent of tactile sensors in robotics has sparked many ideas on how robots can leverage direct contact measurements of their environment interactions to improve manipulation tasks. An important line of research in this regard is that of grasp force control, which aims to manipulate objects safely by limiting the amount of force exerted on the object. While prior works have either hand-modeled their force controllers, employed model-based approaches, or have not shown sim-to-real transfer, we propose a model-free deep reinforcement learning approach trained in simulation and then transferred to the robot without further fine-tuning. We therefore present a simulation environment that produces realistic normal forces, which we use to train continuous force control policies. An evaluation in which we compare against a baseline and perform an ablation study shows that our approach outperforms the hand-modeled baseline and that our proposed inductive bias and domain randomization facilitate sim-to-real transfer. Code, models, and supplementary videos are available on https://sites.google.com/view/rl-force-ctrl

While text style transfer has many applications across natural language processing, the core premise of transferring from a single source style is unrealistic in a real-world setting. In this work, we focus on arbitrary style transfer: rewriting a text from an arbitrary, unknown style to a target style. We propose STEER: Unified Style Transfer with Expert Reinforcement, a unified frame-work developed to overcome the challenge of limited parallel data for style transfer. STEER involves automatically generating a corpus of style-transfer pairs using a product of experts during decoding. The generated offline data is then used to pre-train an initial policy before switching to online, off-policy reinforcement learning for further improvements via fine-grained reward signals. STEER is unified and can transfer to multiple target styles from an arbitrary, unknown source style, making it particularly flexible and efficient. Experimental results on a challenging dataset with text from a diverse set of styles demonstrate state-of-the-art results compared to competitive baselines. Remarkably, STEER outperforms the 175B parameter instruction-tuned GPT-3 on overall style transfer quality, despite being 226 times smaller in size. We also show STEER is robust, maintaining its style transfer capabilities on out-of-domain data, and surpassing nearly all baselines across various styles. The success of our method highlights the potential of RL algorithms when augmented with controllable decoding to overcome the challenge of limited data supervision.

To better match drivers to riders in our ridesharing application, we revised Lyft's core matching algorithm. We use a novel online reinforcement learning approach that estimates the future earnings of drivers in real time and use this information to find more efficient matches. This change was the first documented implementation of a ridesharing matching algorithm that can learn and improve in real time. We evaluated the new approach during weeks of switchback experimentation in most Lyft markets, and estimated how it benefited drivers, riders, and the platform. In particular, it enabled our drivers to serve millions of additional riders each year, leading to more than $30 million per year in incremental revenue. Lyft rolled out the algorithm globally in 2021.

Offline reinforcement learning (RL) offers an appealing approach to real-world tasks by learning policies from pre-collected datasets without interacting with the environment. However, the performance of existing offline RL algorithms heavily depends on the scale and state-action space coverage of datasets. Real-world data collection is often expensive and uncontrollable, leading to small and narrowly covered datasets and posing significant challenges for practical deployments of offline RL. In this paper, we provide a new insight that leveraging the fundamental symmetry of system dynamics can substantially enhance offline RL performance under small datasets. Specifically, we propose a Time-reversal symmetry (T-symmetry) enforced Dynamics Model (TDM), which establishes consistency between a pair of forward and reverse latent dynamics. TDM provides both well-behaved representations for small datasets and a new reliability measure for OOD samples based on compliance with the T-symmetry. These can be readily used to construct a new offline RL algorithm (TSRL) with less conservative policy constraints and a reliable latent space data augmentation procedure. Based on extensive experiments, we find TSRL achieves great performance on small benchmark datasets with as few as 1% of the original samples, which significantly outperforms the recent offline RL algorithms in terms of data efficiency and generalizability. Code is available at: https://github.com/pcheng2/TSRL

We introduce a value-based RL agent, which we 64 call BBF, that achieves super-human performance in the Atari 100K benchmark. BBF relies on scaling 16 the neural networks used for value estimation, as well as a number of other design choices that 4 enable this scaling in a sample-efficient manner. We conduct extensive analyses of these design 1 choices and provide insights for future work. We 2015 2017 2019 2021 2023 end with a discussion about updating the goalposts for sample-efficient RL research on the ALE. Figure 1: Environment samples to reach human-level performance, We make our code and data publicly available.

Automated anesthesia promises to enable more precise and personalized anesthetic administration and free anesthesiologists from repetitive tasks, allowing them to focus on the most critical aspects of a patient's surgical care. Current research has typically focused on creating simulated environments from which agents can learn. These approaches have demonstrated good experimental results, but are still far from clinical application. In this paper, Policy Constraint Q-Learning (PCQL), a data-driven reinforcement learning algorithm for solving the problem of learning anesthesia strategies on real clinical datasets, is proposed. Conservative Q-Learning was first introduced to alleviate the problem of Q function overestimation in an offline context. A policy constraint term is added to agent training to keep the policy distribution of the agent and the anesthesiologist consistent to ensure safer decisions made by the agent in anesthesia scenarios. The effectiveness of PCQL was validated by extensive experiments on a real clinical anesthesia dataset. Experimental results show that PCQL is predicted to achieve higher gains than the baseline approach while maintaining good agreement with the reference dose given by the anesthesiologist, using less total dose, and being more responsive to the patient's vital signs. In addition, the confidence intervals of the agent were investigated, which were able to cover most of the clinical decisions of the anesthesiologist. Finally, an interpretable method, SHAP, was used to analyze the contributing components of the model predictions to increase the transparency of the model.

Applying reinforcement learning (RL) to combinatorial optimization problems is attractive as it removes the need for expert knowledge or pre-solved instances. However, it is unrealistic to expect an agent to solve these (often NP-)hard problems in a single shot at inference due to their inherent complexity. Thus, leading approaches often implement additional search strategies, from stochastic sampling and beam-search to explicit fine-tuning. In this paper, we argue for the benefits of learning a population of complementary policies, which can be simultaneously rolled out at inference. To this end, we introduce Poppy, a simple training procedure for populations. Instead of relying on a predefined or hand-crafted notion of diversity, Poppy induces an unsupervised specialization targeted solely at maximizing the performance of the population. We show that Poppy produces a set of complementary policies, and obtains state-of-the-art RL results on four popular NP-hard problems: traveling salesman, capacitated vehicle routing, 0-1 knapsack, and job-shop scheduling.