Auxiliary tasks have been argued to be useful for representation learning in reinforcement learning. Although many auxiliary tasks have been empirically shown to be effective for accelerating learning on the main task, it is not yet clear what makes useful auxiliary tasks. Some of the most promising results are on the pixel control, reward prediction, and the next state prediction auxiliary tasks; however, the empirical results are mixed, showing substantial improvements in some cases and marginal improvements in others. Careful investigations of how auxiliary tasks help the learning of the main task is necessary. In this paper, we take a step studying the effect of the target policies on the usefulness of the auxiliary tasks formulated as general value functions. General value functions consist of three core elements: 1) policy 2) cumulant 3) continuation function. Our focus on the role of the target policy of the auxiliary tasks is motivated by the fact that the target policy determines the behavior about which the agent wants to make a prediction and the state-action distribution that the agent is trained on, which further affects the main task learning. Our study provides insights about questions such as: Does a greedy policy result in bigger improvement gains compared to other policies? Is it best to set the auxiliary task policy to be the same as the main task policy? Does the choice of the target policy have a substantial effect on the achieved performance gain or simple strategies for setting the policy, such as using a uniformly random policy, work as well? Our empirical results suggest that: 1) Auxiliary tasks with the greedy policy tend to be useful. 2) Most policies, including a uniformly random policy, tend to improve over the baseline. 3) Surprisingly, the main task policy tends to be less useful compared to other policies.
Hierarchical Reinforcement Learning (HRL) is a promising approach to solving long-horizon problems with sparse and delayed rewards. Many existing HRL algorithms either use pre-trained low-level skills that are unadaptable, or require domain-specific information to define low-level rewards. In this paper, we aim to adapt low-level skills to downstream tasks while maintaining the generality of reward design. We propose an HRL framework which sets auxiliary rewards for low-level skill training based on the advantage function of the high-level policy. This auxiliary reward enables efficient, simul…
Off-policy reinforcement learning has many applications including: learning from demonstration, learning multiple goal seeking policies in parallel, and representing predictive knowledge. Recently there has been an proliferation of new policy-evaluation algorithms that fill a longstanding algorithmic void in reinforcement learning: combining robustness to off-policy sampling, function approximation, linear complexity, and temporal difference (TD) updates. This paper contains two main contributions. First, we derive two new hybrid TD policy-evaluation algorithms, which fill a gap in this collec…
University of Alberta（阿尔伯塔大学）；Alberta Machine Intelligence Institute (Amii)（阿尔伯塔机器智能研究所（Amii））
Approximate Policy Iteration (API) algorithms alternate between (approximate) policy evaluation and (approximate) greedification. Many different approaches have been explored for approximate policy evaluation, but less is understood about approximate greedification and what choices guarantee policy improvement. In this work, we investigate approximate greedification when reducing the KL divergence between the parameterized policy and the Boltzmann distribution over action values. In particular, we investigate the difference between the forward and reverse KL divergences, with varying degrees o…
University of Alberta（阿尔伯塔大学）；Alberta Machine Intelligence Institute (Amii)（阿尔伯塔机器智能研究所（Amii））
Learning auxiliary tasks, such as multiple predictions about the world, can provide many benefits to reinforcement learning systems. A variety of off-policy learning algorithms have been developed to learn such predictions, but as yet there is little work on how to adapt the behavior to gather useful data for those off-policy predictions. In this work, we investigate a reinforcement learning system designed to learn a collection of auxiliary tasks, with a behavior policy learning to take actions to improve those auxiliary predictions. We highlight the inherent non-stationarity in this continua…
Predictive auxiliary tasks have been shown to improve performance in numerous reinforcement learning works, however, this effect is still not well understood. The primary purpose of the work presented here is to investigate the impact that an auxiliary task's prediction timescale has on the agent's policy performance. We consider auxiliary tasks which learn to make on-policy predictions using temporal difference learning. We test the impact of prediction timescale using a specific form of auxiliary task in which the input image is used as the prediction target, which we refer to as temporal di…
Recent advances in contextual bandit optimization and reinforcement learning have garnered interest in applying these methods to real-world sequential decision making problems. Real-world applications frequently have constraints with respect to a currently deployed policy. Many of the existing constraint-aware algorithms consider problems with a single objective (the reward) and a constraint on the reward with respect to a baseline policy. However, many important applications involve multiple competing objectives and auxiliary constraints. In this paper, we propose a novel Thompson sampling al…
King’s College London（伦敦国王学院）；University of Oxford（牛津大学）
Off-policy reinforcement learning (RL) from pixel observations is notoriously unstable. As a result, many successful algorithms must combine different domain-specific practices and auxiliary losses to learn meaningful behaviors in complex environments. In this work, we provide novel analysis demonstrating that these instabilities arise from performing temporal-difference learning with a convolutional encoder and low-magnitude rewards. We show that this new visual deadly triad causes unstable training and premature convergence to degenerate solutions, a phenomenon we name catastrophic self-over…
Policy and World Modeling Co-Training for Language Agents
Southern University of Science and Technology（南方科技大学）；Hong Kong University of Science and Technology（香港科学大学）；Hong Kong University of Science and Technology (Guangzhou)（香港科学大学（广州））；Hong Kong Polytechnic University（香港理工大学）；LIGHTSPEED
Standard deep reinforcement learning algorithms use a shared representation for the policy and value function, especially when training directly from images. However, we argue that more information is needed to accurately estimate the value function than to learn the optimal policy. Consequently, the use of a shared representation for the policy and value function can lead to overfitting. To alleviate this problem, we propose two approaches which are combined to create IDAAC: Invariant Decoupled Advantage Actor-Critic. First, IDAAC decouples the optimization of the policy and value function, u…
Shanghai Institute for Advanced Communication and Data Science（上海先进通信与数据科学研究院）；School of Computer Engineering and Science, Shanghai University（上海大学计算机工程与科学学院）
Instability and slowness are two main problems in deep reinforcement learning. Even if proximal policy optimization (PPO) is the state of the art, it still suffers from these two problems. We introduce an improved algorithm based on proximal policy optimization, mixed distributed proximal policy optimization (MDPPO), and show that it can accelerate and stabilize the training process. In our algorithm, multiple different policies train simultaneously and each of them controls several identical agents that interact with environments. Actions are sampled by each policy separately as usual, but th…
In this paper we explore how actor-critic methods in deep reinforcement learning, in particular Asynchronous Advantage Actor-Critic (A3C), can be extended with agent modeling. Inspired by recent works on representation learning and multiagent deep reinforcement learning, we propose two architectures to perform agent modeling: the first one based on parameter sharing, and the second one based on agent policy features. Both architectures aim to learn other agents' policies as auxiliary tasks, besides the standard actor (policy) and critic (values). We performed experiments in both cooperative an…
Google Research（谷歌研究院）；NYU（纽约大学）；University of Alberta（阿尔伯塔大学）
In many real-world applications of reinforcement learning (RL), interactions with the environment are limited due to cost or feasibility. This presents a challenge to traditional RL algorithms since the max-return objective involves an expectation over on-policy samples. We introduce a new formulation of max-return optimization that allows the problem to be re-expressed by an expectation over an arbitrary behavior-agnostic and off-policy data distribution. We first derive this result by considering a regularized version of the dual max-return objective before extending our findings to unregula…
Deterministic Policy Gradient for Learning Equilibrium in Time-Inconsistent Control Problems
Department of Industrial Engineering and Operations Research, University of California, Berkeley, USA（加州大学伯克利分校工业工程与运筹学系）；Department of Applied Mathematics, The Hong Kong Polytechnic University, Kowloon, Hong Kong（香港理工大学应用数学系）
Boosting Reinforcement Learning with Strongly Delayed Feedback Through Auxiliary Short Delays
The University of Liverpool（利物浦大学）；Northwestern University（西北大学）；King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）；National Taiwan University（国立台湾大学）；Nanyang Technological University（南洋理工大学）；The Swiss AI Lab IDSIA/USI/SUPSI（瑞士IDSIA人工智能实验室（隶属于瑞士意大利语区大学与瑞士应用科学大学））；The University of Southampton（南安普顿大学）
University of Twente（特文特大学）；Saxion University of Applied Sciences（萨克逊应用科学大学）
Deep Reinforcement Learning has shown its ability in solving complicated problems directly from high-dimensional observations. However, in end-to-end settings, Reinforcement Learning algorithms are not sample-efficient and requires long training times and quantities of data. In this work, we proposed a framework for sample-efficient Reinforcement Learning that take advantage of state and action representations to transform a high-dimensional problem into a low-dimensional one. Moreover, we seek to find the optimal policy mapping latent states to latent actions. Because now the policy is learne…
We consider the problem of using a heuristic policy to improve the value approximation by the Upper Confidence Bound applied in Trees (UCT) algorithm in non-adversarial settings such as planning with large-state space Markov Decision Processes. Current improvements to UCT focus on either changing the action selection formula at the internal nodes or the rollout policy at the leaf nodes of the search tree. In this work, we propose to add an auxiliary arm to each of the internal nodes, and always use the heuristic policy to roll out simulations at the auxiliary arms. The method aims to get fast …
Health-related data is noisy and stochastic in implying the true physiological states of patients, limiting information contained in single-moment observations for sequential clinical decision making. We model patient-clinician interactions as partially observable Markov decision processes (POMDPs) and optimize sequential treatment based on belief states inferred from history sequence. To facilitate inference, we build a variational generative model and boost state representation with a recurrent neural network (RNN), incorporating an auxiliary loss from sequence auto-encoding. Meanwhile, we o…
University of Amsterdam（阿姆斯特丹大学）；Amazon（亚马逊）；Microsoft Research（微软研究院）
Reinforcement Learning (RL) methods have emerged as a popular choice for training an efficient and effective dialogue policy. However, these methods suffer from sparse and unstable reward signals returned by a user simulator only when a dialogue finishes. Besides, the reward signal is manually designed by human experts, which requires domain knowledge. Recently, a number of adversarial learning methods have been proposed to learn the reward function together with the dialogue policy. However, to alternatively update the dialogue policy and the reward model on the fly, we are limited to policy-…
Faculty of Information Technology（信息技术学院）；Monash University（莫纳什大学）
Scarcity of parallel sentence pairs is a major challenge for training high quality neural machine translation (NMT) models in bilingually low-resource scenarios, as NMT is data-hungry. Multi-task learning is an elegant approach to inject linguistic-related inductive biases into NMT, using auxiliary syntactic and semantic tasks, to improve generalisation. The challenge, however, is to devise effective training schedules, prescribing when to make use of the auxiliary tasks during the training process to fill the knowledge gaps of the main translation task, a setting referred to as biased-MTL. Cu…
Despite success in many challenging problems, reinforcement learning (RL) is still confronted with sample inefficiency, which can be mitigated by introducing prior knowledge to agents. However, many transfer techniques in reinforcement learning make the limiting assumption that the teacher is an expert. In this paper, we use the action prior from the Reinforcement Learning as Inference framework - that is, a distribution over actions at each state which resembles a teacher policy, rather than a Bayesian prior - to recover state-of-the-art policy distillation techniques. Then, we propose a clas…
Tianjin University（天津大学）；College of Intelligence and Computing（智能与计算学部）
Knowing the learning dynamics of policy is significant to unveiling the mysteries of Reinforcement Learning (RL). It is especially crucial yet challenging to Deep RL, from which the remedies to notorious issues like sample inefficiency and learning instability could be obtained. In this paper, we study how the policy networks of typical DRL agents evolve during the learning process by empirically investigating several kinds of temporal change for each policy parameter. On typical MuJoCo and DeepMind Control Suite (DMC) benchmarks, we find common phenomena for TD3 and RAD agents: 1) the activit…
Department of Computer Science, University of California, Los Angeles, CA, USA（加州大学洛杉矶分校计算机科学系）；Princeton AI Laboratory, Princeton University, Princeton, NJ, USA（普林斯顿大学普林斯顿AI实验室）
University of Electronic Science and Technology of China（电子科技大学）；Institute for Infocomm Research, Agency for Science, Technology and Research（新加坡科技研究局信息通信研究院）
This paper investigates a master unmanned aerial vehicle (MUAV)-powered Internet of Things (IoT) network, in which we propose using a rechargeable auxiliary UAV (AUAV) equipped with an intelligent reflecting surface (IRS) to enhance the communication signals from the MUAV and also leverage the MUAV as a recharging power source. Under the proposed model, we investigate the optimal collaboration strategy of these energy-limited UAVs to maximize the accumulated throughput of the IoT network. Depending on whether there is charging between the two UAVs, two optimization problems are formulated. To …
University of Alberta（阿尔伯塔大学）；Stanford University（斯坦福大学）；University of Massachusetts Amherst（马萨诸塞大学阿默斯特分校）
In this work, we consider the off-policy policy evaluation problem for contextual bandits and finite horizon reinforcement learning in the nonstationary setting. Reusing old data is critical for policy evaluation, but existing estimators that reuse old data introduce large bias such that we can not obtain a valid confidence interval. Inspired from a related field called survey sampling, we introduce a variant of the doubly robust (DR) estimator, called the regression-assisted DR estimator, that can incorporate the past data without introducing a large bias. The estimator unifies several existi…
RLFR: Extending Reinforcement Learning for LLMs with Flow Environment
University of Science and Technology of China（中国科学技术大学）；Shanghai Innovation Institute（上海创新研究院）；ByteDance（字节跳动）；Wuhan University（武汉大学）；Southeast University（东南大学）
Reinforcement learning with function approximation can be unstable and even divergent, especially when combined with off-policy learning and Bellman updates. In deep reinforcement learning, these issues have been dealt with empirically by adapting and regularizing the representation, in particular with auxiliary tasks. This suggests that representation learning may provide a means to guarantee stability. In this paper, we formally show that there are indeed nontrivial state representations under which the canonical TD algorithm is stable, even when learning off-policy. We analyze representatio…
Sequential experiments are often characterized by an exploration-exploitation tradeoff that is captured by the multi-armed bandit (MAB) framework. This framework has been studied and applied, typically when at each time period feedback is received only on the action that was selected at that period. However, in many practical settings additional data may become available between decision epochs. We introduce a generalized MAB formulation, which considers a broad class of distributions that are informative about mean rewards, and allows observations from these distributions to arrive according …
Maluuba, A Microsoft Company（Maluuba，微软旗下公司）；McGill University（麦吉尔大学）
Policy gradient methods in reinforcement learning have become increasingly prevalent for state-of-the-art performance in continuous control tasks. Novel methods typically benchmark against a few key algorithms such as deep deterministic policy gradients and trust region policy optimization. As such, it is important to present and use consistent baselines experiments. However, this can be difficult due to general variance in the algorithms, hyper-parameter tuning, and environment stochasticity. We investigate and discuss: the significance of hyper-parameters in policy gradients for continuous c…
University of Oxford（牛津大学）；Imperial College London（帝国理工学院）
Using privileged information during training can improve the sample efficiency and performance of machine learning systems. This paradigm has been applied to reinforcement learning (RL), primarily in the form of distillation or auxiliary tasks, and less commonly in the form of augmenting the inputs of agents. In this work, we investigate Privileged Information Dropout (\pid) for achieving the latter which can be applied equally to value-based and policy-based RL algorithms. Within a simple partially-observed environment, we demonstrate that \pid outperforms alternatives for leveraging privileg…
From $P(y|x)$ to $P(y)$: Investigating Reinforcement Learning in Pre-train Space
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；University of Chinese Academy of Sciences（中国科学院大学）；National University of Singapore（新加坡国立大学）；Tencent AI Lab（腾讯AI实验室）
University of Science and Technology of China（中国科学技术大学）；University of New South Wales（新南威尔士大学）
Due to its property of not requiring prior knowledge of the environment, reinforcement learning has significant potential for quantum control problems. In this work, we investigate the effectiveness of continuous control policies based on deep deterministic policy gradient. To solve the sparse reward signal in quantum learning control problems, we propose an auxiliary task-based deep reinforcement learning (AT-DRL) for quantum control. In particular, we first design a guided reward function based on the fidelity of quantum states that enables incremental fidelity improvement. Then, we introduc…
We study the roots of algorithmic progress in deep policy gradient algorithms through a case study on two popular algorithms: Proximal Policy Optimization (PPO) and Trust Region Policy Optimization (TRPO). Specifically, we investigate the consequences of "code-level optimizations:" algorithm augmentations found only in implementations or described as auxiliary details to the core algorithm. Seemingly of secondary importance, such optimizations turn out to have a major impact on agent behavior. Our results show that they (a) are responsible for most of PPO's gain in cumulative reward over TRPO,…
