One of the remaining challenges in reinforcement learning is to develop agents that can generalise to novel scenarios they might encounter once deployed. This challenge is often framed in a multi-task setting where agents train on a fixed set of tasks and have to generalise to new tasks. Recent work has shown that in this setting increased exploration during training can be leveraged to increase the generalisation performance of the agent. This makes sense when the states encountered during testing can actually be explored during training. In this paper, we provide intuition why exploration can also benefit generalisation to states that cannot be explicitly encountered during training. Additionally, we propose a novel method Explore-Go that exploits this intuition by increasing the number of states on which the agent trains. Explore-Go effectively increases the starting state distribution of the agent and as a result can be used in conjunction with most existing on-policy or off-policy reinforcement learning algorithms. We show empirically that our method can increase generalisation performance in an illustrative environment and on the Procgen benchmark.
We introduce a new count-based optimistic exploration algorithm for Reinforcement Learning (RL) that is feasible in environments with high-dimensional state-action spaces. The success of RL algorithms in these domains depends crucially on generalisation from limited training experience. Function approximation techniques enable RL agents to generalise in order to estimate the value of unvisited states, but at present few methods enable generalisation regarding uncertainty. This has prevented the combination of scalable RL algorithms with efficient exploration strategies that drive the agent to …
This paper introduces a method for constructing an upper bound for exploration policy using either the weighted variance of return sequences or the weighted temporal difference (TD) error. We demonstrate that the variance of the return sequence for a specific state-action pair is an important information source that can be leveraged to guide exploration in reinforcement learning. The intuition is that fluctuation in the return sequence indicates greater uncertainty in the near future returns. This divergence occurs because of the cyclic nature of value-based reinforcement learning; the evolvin…
While in general trading off exploration and exploitation in reinforcement learning is hard, under some formulations relatively simple solutions exist. In this paper, we first derive upper bounds for the utility of selecting different actions in the multi-armed bandit setting. Unlike the common statistical upper confidence bounds, these explicitly link the planning horizon, uncertainty and the need for exploration explicit. The resulting algorithm can be seen as a generalisation of the classical Thompson sampling algorithm. We experimentally test these algorithms, as well as $ε$-greedy and the…
Many real world problems can be expressed as optimisation problems. Solving this kind of problems means to find, among all possible solutions, the one that maximises an evaluation function. One approach to solve this kind of problem is to use an informed search strategy. The principle of this kind of strategy is to use problem-specific knowledge beyond the definition of the problem itself to find solutions more efficiently than with an uninformed strategy. This kind of strategy demands to define problem-specific knowledge (heuristics). The efficiency and the effectiveness of systems based on i…
University of Amsterdam（阿姆斯特丹大学）；Vrije Universiteit Amsterdam（阿姆斯特丹自由大学）
Enabling reinforcement learning (RL) agents to leverage a knowledge base while learning from experience promises to advance RL in knowledge intensive domains. However, it has proven difficult to leverage knowledge that is not manually tailored to the environment. We propose to use the subclass relationships present in open-source knowledge graphs to abstract away from specific objects. We develop a residual policy gradient method that is able to integrate knowledge across different abstraction levels in the class hierarchy. Our method results in improved sample efficiency and generalisation to…
Directed exploration strategies for reinforcement learning are critical for learning an optimal policy in a minimal number of interactions with the environment. Many algorithms use optimism to direct exploration, either through visitation estimates or upper confidence bounds, as opposed to data-inefficient strategies like ε-greedy that use random, undirected exploration. Most data-efficient exploration methods require significant computation, typically relying on a learned model to guide exploration. Least-squares methods have the potential to provide some of the data-efficiency benefits of mo…
Leveraging Temporally Extended Behavior Sharing for Multi-task Reinforcement Learning
Department of Aerospace Engineering, Seoul National University（首尔国立大学航空航天工程系）；Artificial Intelligence Institute, Seoul National University（首尔国立大学人工智能研究所）
Dual-Flow Reinforcement Learning with State-Aware Exploration
State Key Laboratory of Intelligent Green Vehicle and Mobility, Tsinghua University（智能绿色车辆与移动国家重点实验室，清华大学）；Tsinghua University-Toyota Joint Center（清华大学-丰田联合中心）；Tsinghua University–SAIC GM Wuling Joint Research Center（清华大学-上汽通用五菱联合研究中心）
The goal of meta-reinforcement learning (meta-RL) is to build agents that can quickly learn new tasks by leveraging prior experience on related tasks. Learning a new task often requires both exploring to gather task-relevant information and exploiting this information to solve the task. In principle, optimal exploration and exploitation can be learned end-to-end by simply maximizing task performance. However, such meta-RL approaches struggle with local optima due to a chicken-and-egg problem: learning to explore requires good exploitation to gauge the exploration's utility, but learning to exp…
$β$-DQN: Improving Deep Q-Learning By Evolving the Behavior
University of Alberta and Amii（阿尔伯塔大学和Amii）；Shanghai Jiao Tong University（上海交通大学）；Zhongguancun Academy（中关村学院）；Edmonton Research Center, Huawei（埃德蒙顿研究中心，华为）
Exploration is a key challenge in Reinforcement Learning, especially in long-horizon, deceptive and sparse-reward environments. For such applications, population-based approaches have proven effective. Methods such as Quality-Diversity deals with this by encouraging novel solutions and producing a diversity of behaviours. However, these methods are driven by either undirected sampling (i.e. mutations) or use approximated gradients (i.e. Evolution Strategies) in the parameter space, which makes them highly sample-inefficient. In this paper, we propose a model-based Quality-Diversity approach. I…
Humans learn to play video games significantly faster than the state-of-the-art reinforcement learning (RL) algorithms. People seem to build simple models that are easy to learn to support planning and strategic exploration. Inspired by this, we investigate two issues in leveraging model-based RL for sample efficiency. First we investigate how to perform strategic exploration when exact planning is not feasible and empirically show that optimistic Monte Carlo Tree Search outperforms posterior sampling methods. Second we show how to learn simple deterministic models to support fast learning usi…
University College London（伦敦大学学院）；Inria（法国国家信息与自动化研究所）
We introduce a modified version of the excess risk, which can be used to obtain tighter, fast-rate PAC-Bayesian generalisation bounds. This modified excess risk leverages information about the relative hardness of data examples to reduce the variance of its empirical counterpart, tightening the bound. We combine this with a new bound for $[-1, 1]$-valued (and potentially non-independent) signed losses, which is more favourable when they empirically have low variance around $0$. The primary new technical tool is a novel result for sequences of interdependent random vectors which may be of indep…
Test-Time Training Scaling Laws for Chemical Exploration in Drug Design
Computational Science Laboratory, Universitat Pompeu Fabra, Barcelona Biomedical Research Park (PRBB), C Dr. Aiguader 88, 08003 Barcelona, Spain（计算科学实验室，庞培法布拉大学，巴塞罗那生物医学研究公园（PRBB），C Dr. Aiguader 88，08003巴塞罗那，西班牙）
Peking University（北京大学）；University of Minnesota（明尼苏达大学）；Microsoft Research（微软研究院）；Carnegie Mellon University（卡内基梅隆大学）
Consider a bandit algorithm that recommends actions to self-interested users in a recommendation system. The users are free to choose other actions and need to be incentivized to follow the algorithm's recommendations. While the users prefer to exploit, the algorithm can incentivize them to explore by leveraging the information collected from the previous users. All published work on this problem, known as incentivized exploration, focuses on small, unstructured action sets and mainly targets the case when the users' beliefs are independent across actions. However, realistic exploration proble…
Carnegie Mellon University（卡内基梅隆大学）；Bosch Research Pittsburgh（博世匹兹堡研究院）；NVIDIA（英伟达）；AICrowd；Amazon（亚马逊）；Arrival Ltd.（Arrival公司）；Pinterest
We present the results of our autonomous racing virtual challenge, based on the newly-released Learn-to-Race (L2R) simulation framework, which seeks to encourage interdisciplinary research in autonomous driving and to help advance the state of the art on a realistic benchmark. Analogous to racing being used to test cutting-edge vehicles, we envision autonomous racing to serve as a particularly challenging proving ground for autonomous agents as: (i) they need to make sub-second, safety-critical decisions in a complex, fast-changing environment; and (ii) both perception and control must be robu…
Microsoft Research（微软研究院）；Harvard University（哈佛大学）
Intrinsic rewards play a central role in handling the exploration-exploitation trade-off when designing sequential decision-making algorithms, in both foundational theory and state-of-the-art deep reinforcement learning. The LinUCB algorithm, a centerpiece of the stochastic linear bandits literature, prescribes an elliptical bonus which addresses the challenge of leveraging shared information in large action spaces. This bonus scheme cannot be directly transferred to high-dimensional exploration problems, however, due to the computational cost of maintaining the inverse covariance matrix of ac…
National University of Defense Technology（中国人民解放军国防科技大学）
This paper presents a time-efficient scheme for Mars exploration by the cooperation of multiple drones and a rover. To maximize effective coverage of the Mars surface in the long run, a comprehensive framework has been developed with joint consideration for limited energy, sensor model, communication range and safety radius, which we call TIME-SC2 (TIme-efficient Mars Exploration of Simultaneous Coverage and Charging). First, we propose a multi-drone coverage control algorithm by leveraging emerging deep reinforcement learning and design a novel information map to represent dynamic system stat…
Intrinsic-Motivation Multi-Robot Social Formation Navigation with Coordinated Exploration
School of Computer Science and Technology（计算机科学与技术学院）；Wuhan University of Science and Technology（武汉科技大学）；Hubei Province Key Laboratory of Intelligent Information Processing（湖北省智能信息处理重点实验室）
This paper considers stochastic bandits with side observations, a model that accounts for both the exploration/exploitation dilemma and relationships between arms. In this setting, after pulling an arm i, the decision maker also observes the rewards for some other actions related to i. We will see that this model is suited to content recommendation in social networks, where users' reactions may be endorsed or not by their friends. We provide efficient algorithms based on upper confidence bounds (UCBs) to leverage this additional information and derive new bounds improving on standard regret gu…
University of Pennsylvania（宾夕法尼亚大学）；UC Berkeley（加州大学伯克利分校）；Google（谷歌）；Facebook AI Research（Facebook AI研究院）
Reinforcement learning allows solving complex tasks, however, the learning tends to be task-specific and the sample efficiency remains a challenge. We present Plan2Explore, a self-supervised reinforcement learning agent that tackles both these challenges through a new approach to self-supervised exploration and fast adaptation to new tasks, which need not be known during exploration. During exploration, unlike prior methods which retrospectively compute the novelty of observations after the agent has already reached them, our agent acts efficiently by leveraging planning to seek out expected f…
The ability of robots to grasp novel objects has industry applications in e-commerce order fulfillment and home service. Data-driven grasping policies have achieved success in learning general strategies for grasping arbitrary objects. However, these approaches can fail to grasp objects which have complex geometry or are significantly outside of the training distribution. We present a Thompson sampling algorithm that learns to grasp a given object with unknown geometry using online experience. The algorithm leverages learned priors from the Dexterity Network robot grasp planner to guide grasp …
McGill University（麦吉尔大学）；Mila -- Quebec AI Institute（米拉-魁北克人工智能研究所）；Université Laval（拉瓦尔大学）；Facebook AI Research（脸书人工智能研究院）；Microsoft Research Montreal（微软研究院蒙特利尔分院）；Université de Montréal（蒙特利尔大学）
The ability to discover approximately optimal policies in domains with sparse rewards is crucial to applying reinforcement learning (RL) in many real-world scenarios. Approaches such as neural density models and continuous exploration (e.g., Go-Explore) have been proposed to maintain the high exploration rate necessary to find high performing and generalizable policies. Soft actor-critic(SAC) is another method for improving exploration that aims to combine efficient learning via off-policy updates while maximizing the policy entropy. In this work, we extend SAC to a richer class of probability…
Designing agent that can autonomously discover and learn a diversity of structures and skills in unknown changing environments is key for lifelong machine learning. A central challenge is how to learn incrementally representations in order to progressively build a map of the discovered structures and re-use it to further explore. To address this challenge, we identify and target several key functionalities. First, we aim to build lasting representations and avoid catastrophic forgetting throughout the exploration process. Secondly we aim to learn a diversity of representations allowing to disc…
RoboEXP: Action-Conditioned Scene Graph via Interactive Exploration for Robotic Manipulation
Peking University（北京大学）；National University of Singapore（新加坡国立大学）；Columbia University（哥伦比亚大学）；University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Amazon（亚马逊公司）
