The objective of lifelong reinforcement learning (RL) is to optimize agents which can continuously adapt and interact in changing environments. However, current RL approaches fail drastically when environments are non-stationary and interactions are non-episodic. We propose Lifelong Skill Planning (LiSP), an algorithmic framework for non-episodic lifelong RL based on planning in an abstract space of higher-order skills. We learn the skills in an unsupervised manner using intrinsic rewards and plan over the learned skills using a learned dynamics model. Moreover, our framework permits skill discovery even from offline data, thereby reducing the need for excessive real-world interactions. We demonstrate empirically that LiSP successfully enables long-horizon planning and learns agents that can avoid catastrophic failures even in challenging non-stationary and non-episodic environments derived from gridworld and MuJoCo benchmarks.
Learning While Acting: A Skill-Enhanced Test-Time Co-Evolution Framework for Online Lifelong Learning Agents
School of Computer Science and Technology, East China Normal University（东华大学计算机科学与技术学院）；Shanghai AI Laboratory（上海人工智能实验室）；Software Engineering Institute, East China Normal University（东华大学软件工程学院）
We propose a lifelong learning system that has the ability to reuse and transfer knowledge from one task to another while efficiently retaining the previously learned knowledge-base. Knowledge is transferred by learning reusable skills to solve tasks in Minecraft, a popular video game which is an unsolved and high-dimensional lifelong learning problem. These reusable skills, which we refer to as Deep Skill Networks, are then incorporated into our novel Hierarchical Deep Reinforcement Learning Network (H-DRLN) architecture using two techniques: (1) a deep skill array and (2) skill distillation,…
We introduce the eigentask framework for lifelong learning. An eigentask is a pairing of a skill that solves a set of related tasks, paired with a generative model that can sample from the skill's input space. The framework extends generative replay approaches, which have mainly been used to avoid catastrophic forgetting, to also address other lifelong learning goals such as forward knowledge transfer. We propose a wake-sleep cycle of alternating task learning and knowledge consolidation for learning in our framework, and instantiate it for lifelong supervised learning and lifelong RL. We achi…
Reinforcement learning has the potential to automate the acquisition of behavior in complex settings, but in order for it to be successfully deployed, a number of practical challenges must be addressed. First, in real world settings, when an agent attempts a task and fails, the environment must somehow "reset" so that the agent can attempt the task again. While easy in simulation, this could require considerable human effort in the real world, especially if the number of trials is very large. Second, real world learning often involves complex, temporally extended behavior that is often difficu…
Multi-task learning ideally allows robots to acquire a diverse repertoire of useful skills. However, many multi-task reinforcement learning efforts assume the robot can collect data from all tasks at all times. In reality, the tasks that the robot learns arrive sequentially, depending on the user and the robot's current environment. In this work, we study a practical sequential multi-task RL problem that is motivated by the practical constraints of physical robotic systems, and derive an approach that effectively leverages the data and policies learned for previous tasks to cumulatively grow t…
Santa Fe Institute（圣塔菲研究所）；University of California, Merced（加州大学默塞德分校）
Lifelong learning occurs on timescales ranging from minutes to decades. People can lose themselves in a new skill, practicing for hours until exhausted. And they can pursue mastery over days or decades, perhaps abandoning old skills entirely to seek out new challenges. A full understanding of learning requires an account that integrates these timescales. Here, we present a minimal quantitative model that unifies the nested timescales of learning. Our dynamical model recovers classic accounts of skill acquisition, and describes how learning emerges from moment-to-moment dynamics of motivation, …
Johannes Kepler University Linz（林茨约翰内斯·开普勒大学）；Dynatrace Research（Dynatrace研究院）；Institute of Advanced Research in Artificial Intelligence (IARAI)（高级人工智能研究所（IARAI））
In lifelong learning, an agent learns throughout its entire life without resets, in a constantly changing environment, as we humans do. Consequently, lifelong learning comes with a plethora of research problems such as continual domain shifts, which result in non-stationary rewards and environment dynamics. These non-stationarities are difficult to detect and cope with due to their continuous nature. Therefore, exploration strategies and learning methods are required that are capable of tracking the steady domain shifts, and adapting to them. We propose Reactive Exploration to track and react …
We study learning control in an online reset-free lifelong learning scenario, where mistakes can compound catastrophically into the future and the underlying dynamics of the environment may change. Traditional model-free policy learning methods have achieved successes in difficult tasks due to their broad flexibility, but struggle in this setting, as they can activate failure modes early in their lifetimes which are difficult to recover from and face performance degradation as dynamics change. On the other hand, model-based planning methods learn and adapt quickly, but require prohibitive leve…
Mila（米拉研究所）；McGill University（麦吉尔大学）；Université de Montréal（蒙特利尔大学）
To achieve general artificial intelligence, reinforcement learning (RL) agents should learn not only to optimize returns for one specific task but also to constantly build more complex skills and scaffold their knowledge about the world, without forgetting what has already been learned. In this paper, we discuss the desired characteristics of environments that can support the training and evaluation of lifelong reinforcement learning agents, review existing environments from this perspective, and propose recommendations for devising suitable environments in the future.
Universität Hamburg（汉堡大学）；Rochester Institute of Technology（罗切斯特理工学院）；Heriot-Watt University（赫瑞-瓦特大学）
Humans and animals have the ability to continually acquire, fine-tune, and transfer knowledge and skills throughout their lifespan. This ability, referred to as lifelong learning, is mediated by a rich set of neurocognitive mechanisms that together contribute to the development and specialization of our sensorimotor skills as well as to long-term memory consolidation and retrieval. Consequently, lifelong learning capabilities are crucial for autonomous agents interacting in the real world and processing continuous streams of information. However, lifelong learning remains a long-standing chall…
Shared autonomy enables robots to infer user intent and assist in accomplishing it. But when the user wants to do a new task that the robot does not know about, shared autonomy will hinder their performance by attempting to assist them with something that is not their intent. Our key idea is that the robot can detect when its repertoire of intents is insufficient to explain the user's input, and give them back control. This then enables the robot to observe unhindered task execution, learn the new intent behind it, and add it to this repertoire. We demonstrate with both a case study and a user…
The Technion - Israel Institute of Technology（以色列理工学院）；Google Deepmind（谷歌DeepMind）
We introduce the Adaptive Skills, Adaptive Partitions (ASAP) framework that (1) learns skills (i.e., temporally extended actions or options) as well as (2) where to apply them. We believe that both (1) and (2) are necessary for a truly general skill learning framework, which is a key building block needed to scale up to lifelong learning agents. The ASAP framework can also solve related new tasks simply by adapting where it applies its existing learned skills. We prove that ASAP converges to a local optimum under natural conditions. Finally, our experimental results, which include a RoboCup do…
Humans can learn a variety of concepts and skills incrementally over the course of their lives while exhibiting many desirable properties, such as continual learning without forgetting, forward transfer and backward transfer of knowledge, and learning a new concept or task with only a few examples. Several lines of machine learning research, such as lifelong learning, few-shot learning, and transfer learning, attempt to capture these properties. However, most previous approaches can only demonstrate subsets of these properties, often by different complex mechanisms. In this work, we propose a …
University of Oxford（牛津大学）；Facebook AI Research（脸书人工智能研究院）
In lifelong learning, the learner is presented with a sequence of tasks, incrementally building a data-driven prior which may be leveraged to speed up learning of a new task. In this work, we investigate the efficiency of current lifelong approaches, in terms of sample complexity, computational and memory cost. Towards this end, we first introduce a new and a more realistic evaluation protocol, whereby learners observe each example only once and hyper-parameter selection is done on a small and disjoint set of tasks, which is not used for the actual learning experience and evaluation. Second, w…
T2S: Tokenized Skill Scaling for Lifelong Imitation Learning
School of Computer Science and Technology, East China Normal University, Shanghai 200062, China（东华大学计算机科学与技术学院）；Shanghai Innovation Institute, Shanghai 200062, China（上海创新研究院）；School of Informatics, Xiamen University, Xiamen, Fujian 361005, China（厦门大学信息学院）
Human beings are able to master a variety of knowledge and skills with ongoing learning. By contrast, dramatic performance degradation is observed when new tasks are added to an existing neural network model. This phenomenon, termed as \emph{Catastrophic Forgetting}, is one of the major roadblocks that prevent deep neural networks from achieving human-level artificial intelligence. Several research efforts, e.g. \emph{Lifelong} or \emph{Continual} learning algorithms, have been proposed to tackle this problem. However, they either suffer from an accumulating drop in performance as the task seq…
ENSTA Paris（巴黎高等先进技术学院）；Institut Polytechnique de Paris（巴黎理工学院）
Despite the recent progress in deep learning and reinforcement learning, transfer and generalization of skills learned on specific tasks is very limited compared to human (or animal) intelligence. The lifelong, incremental building of common sense knowledge might be a necessary component on the way to achieve more general intelligence. A promising direction is to build world models capturing the true physical mechanisms hidden behind the sensorimotor interaction with the environment. Here we explore the idea that inferring the causal structure of the environment could benefit from well-chosen …
Demonstration-free Autonomous Reinforcement Learning via Implicit and Bidirectional Curriculum
Seoul National University（首尔国立大学）；Artificial Intelligence Institute of Seoul National University (AIIS)（首尔国立大学人工智能研究所（AIIS））；Automation and Systems Research Institute (ASRI)（自动化与系统研究所（ASRI））
Reinforcement learning (RL) algorithms typically start tabula rasa, without any prior knowledge of the environment, and without any prior skills. This however often leads to low sample efficiency, requiring a large amount of interaction with the environment. This is especially true in a lifelong learning setting, in which the agent needs to continually extend its capabilities. In this paper, we examine how a pre-trained task-independent language model can make a goal-conditional RL agent more sample efficient. We do this by facilitating transfer learning between different related tasks. We exp…
University of Washington（华盛顿大学）；University of California, Berkeley（加州大学伯克利分校）
Autonomous learning of robotic skills can allow general-purpose robots to learn wide behavioral repertoires without requiring extensive manual engineering. However, robotic skill learning methods typically make one of several trade-offs to enable practical real-world learning, such as requiring manually designed policy or value function representations, initialization from human-provided demonstrations, instrumentation of the training environment, or extremely long training times. In this paper, we propose a new reinforcement learning algorithm for learning manipulation skills that can train g…
UC Berkeley（加州大学伯克利分校）；University of Washington（华盛顿大学）
Reinforcement Learning (RL) algorithms can in principle acquire complex robotic skills by learning from large amounts of data in the real world, collected via trial and error. However, most RL algorithms use a carefully engineered setup in order to collect data, requiring human supervision and intervention to provide episodic resets. This is particularly evident in challenging robotics problems, such as dexterous manipulation. To make data collection scalable, such applications require reset-free algorithms that are able to learn autonomously, without explicit instrumentation or human interven…
ETH Zürich（苏黎世联邦理工学院）；University of Rochester（罗切斯特大学）；Carnegie Mellon University（卡内基梅隆大学）；Paris Artificial Intelligence Research Institute, Inria（法国国家信息与自动化研究所巴黎人工智能研究院）
Curiosity is a vital metacognitive skill in educational contexts, leading to creativity, and a love of learning. And while many school systems increasingly undercut curiosity by teaching to the test, teachers are increasingly interested in how to evoke curiosity in their students to prepare them for a world in which lifelong learning and reskilling will be more and more important. One aspect of curiosity that has received little attention, however, is the role of peers in eliciting curiosity. We present what we believe to be the first theoretical framework that articulates an integrated socio-…
To be capable of lifelong learning in a real-life environment, robots have to tackle multiple challenges. Being able to relate physical properties they may observe in their environment to possible interactions they may have is one of them. This skill, named affordance learning, is strongly related to embodiment and is mastered through each person's development: each individual learns affordances differently through their own interactions with their surroundings. Current methods for affordance learning usually use either fixed actions to learn these affordances or focus on static setups involvi…
Continual Policy Consolidation for Lifelong Robot Learning
Department of XXX, University of YYY, Location, Country；School of ZZZ, Institute of WWW, Location, Country；Zhejiang University, Zhejiang, China；South University of Science；Hong Kong University, Hong Kong, China；INFIFORCE Intelligent Technology Co., Ltd., Hangzhou, China；The Hong Kong Polytechnic University, Hong Kong, China；School of Computer Science, Carnegie Mellon University
Google Brain（谷歌大脑）；University of Cambridge（剑桥大学）；Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；UC Berkeley（加州大学伯克利分校）
Deep reinforcement learning algorithms can learn complex behavioral skills, but real-world application of these methods requires a large amount of experience to be collected by the agent. In practical settings, such as robotics, this involves repeatedly attempting a task, resetting the environment between each attempt. However, not all tasks are easily or automatically reversible. In practice, this learning process requires extensive human intervention. In this work, we propose an autonomous method for safe and efficient reinforcement learning that simultaneously learns a forward and reset pol…
Seoul National University（首尔国立大学）；Automation and Systems Research Institute (ASRI)（自动化与系统研究所）
Deep reinforcement learning has enabled robots to learn motor skills from environmental interactions with minimal to no prior knowledge. However, existing reinforcement learning algorithms assume an episodic setting, in which the agent resets to a fixed initial state distribution at the end of each episode, to successfully train the agents from repeated trials. Such reset mechanism, while trivial for simulated tasks, can be challenging to provide for real-world robotics tasks. Resets in robotic systems often require extensive human supervision and task-specific workarounds, which contradicts t…
Quality-Diversity (QD) algorithms can discover large and complex behavioural repertoires consisting of both diverse and high-performing skills. However, the generation of behavioural repertoires has mainly been limited to simulation environments instead of real-world learning. This is because existing QD algorithms need large numbers of evaluations as well as episodic resets, which require manual human supervision and interventions. This paper proposes Reset-Free Quality-Diversity optimization (RF-QD) as a step towards autonomous learning for robotics in open-ended environments. We build on Dy…
Ritsumeikan University（立命馆大学）；Keio University（庆应义塾大学）；The University of Tokyo（东京大学）；Università Milano-Bicocca（米兰比可卡大学）；University of Essex（埃塞克斯大学）；Donders Institute for Brain, Cognition and Behaviour（唐德斯脑、认知与行为研究所）；Cajal International Neuroscience Center, Spanish National Research Council（西班牙国家研究委员会卡哈尔国际神经科学中心）；Bogazici University（博阿齐奇大学）；Queen Mary University of London（伦敦玛丽女王大学）；The University of Electro-Communications（电气通信大学）；National Autonomous University of Mexico（墨西哥国立自治大学）；Universidad Autónoma del Estado de Morelos（莫雷洛斯州自治大学）；Institute of Cognitive Sciences and Technologies, National Research Council of Italy（意大利国家研究委员会认知科学与技术研究所）
Creating autonomous robots that can actively explore the environment, acquire knowledge and learn skills continuously is the ultimate achievement envisioned in cognitive and developmental robotics. Their learning processes should be based on interactions with their physical and social world in the manner of human learning and cognitive development. Based on this context, in this paper, we focus on the two concepts of world models and predictive coding. Recently, world models have attracted renewed attention as a topic of considerable interest in artificial intelligence. Cognitive systems learn…
The automatic and efficient discovery of skills, without supervision, for long-living autonomous agents, remains a challenge of Artificial Intelligence. Intrinsically Motivated Goal Exploration Processes give learning agents a human-inspired mechanism to sequentially select goals to achieve. This approach gives a new perspective on the lifelong learning problem, with promising results on both simulated and real-world experiments. Until recently, those algorithms were restricted to domains with experimenter-knowledge, since the Goal Space used by the agents was built on engineered feature extra…
Reinforcement learning (RL) algorithms hold the promise of enabling autonomous skill acquisition for robotic systems. However, in practice, real-world robotic RL typically requires time consuming data collection and frequent human intervention to reset the environment. Moreover, robotic policies learned with RL often fail when deployed beyond the carefully controlled setting in which they were learned. In this work, we study how these challenges can all be tackled by effective utilization of diverse offline datasets collected from previously seen tasks. When faced with a new task, our system a…
A Definition of Open-Ended Learning Problems for Goal-Conditioned Agents
Sorbonne Université（索邦大学）；CNRS（法国国家科学研究中心）；Institut des Systèmes Intelligents et de Robotique（智能系统与机器人研究所）；INRIA Bordeaux Sud-Ouest（法国国家信息与自动化研究所波尔多南部-西部分部）；Institute of Cognitive Sciences and Technologies（认知科学与技术研究所）；CNR（意大利国家研究委员会）；CITIC（信息与通信技术中心）；Universidade da Coruña（拉科鲁尼亚大学）；Massachusetts Institute of Technology（麻省理工学院）
Robot learning is a very promising topic for the future of automation and machine intelligence. Future robots should be able to autonomously acquire skills, learn to represent their environment, and interact with it. While these topics have been explored in simulation, real-world robot learning research seems to be still limited. This is due to the additional challenges encountered in the real-world, such as noisy sensors and actuators, safe exploration, non-stationary dynamics, autonomous environment resetting as well as the cost of running experiments for long periods of time. Unless we deve…
University of the Witwatersrand（威特沃特斯兰德大学）；Council for Scientific and Industrial Research（科学与工业研究委员会）
An important property for lifelong-learning agents is the ability to combine existing skills to solve unseen tasks. In general, however, it is unclear how to compose skills in a principled way. We provide a "recipe" for optimal value function composition in entropy-regularised reinforcement learning (RL) and then extend this to the standard RL setting. Composition is demonstrated in a video game environment, where an agent with an existing library of policies is able to solve new tasks without the need for further learning.
School of Computer Science and Applied Mathematics（计算机科学与应用数学系）；University of the Witwatersrand（金山大学）
The ability to compose learned skills to solve new tasks is an important property of lifelong-learning agents. In this work, we formalise the logical composition of tasks as a Boolean algebra. This allows us to formulate new tasks in terms of the negation, disjunction and conjunction of a set of base tasks. We then show that by learning goal-oriented value functions and restricting the transition dynamics of the tasks, an agent can solve these new tasks with no further learning. We prove that by composing these value functions in specific ways, we immediately recover the optimal policies for a…
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…
