Learning effective representations in image-based environments is crucial for sample efficient Reinforcement Learning (RL). Unfortunately, in RL, representation learning is confounded with the exploratory experience of the agent -- learning a useful representation requires diverse data, while effective exploration is only possible with coherent representations. Furthermore, we would like to learn representations that not only generalize across tasks but also accelerate downstream exploration for efficient task-specific training. To address these challenges we propose Proto-RL, a self-supervised framework that ties representation learning with exploration through prototypical representations. These prototypes simultaneously serve as a summarization of the exploratory experience of an agent as well as a basis for representing observations. We pre-train these task-agnostic representations and prototypes on environments without downstream task information. This enables state-of-the-art downstream policy learning on a set of difficult continuous control tasks.
Top-performing Model-Based Reinforcement Learning (MBRL) agents, such as Dreamer, learn the world model by reconstructing the image observations. Hence, they often fail to discard task-irrelevant details and struggle to handle visual distractions. To address this issue, previous work has proposed to contrastively learn the world model, but the performance tends to be inferior in the absence of distractions. In this paper, we seek to enhance robustness to distractions for MBRL agents. Specifically, we consider incorporating prototypical representations, which have yielded more accurate and robu…
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）；The University of Hong Kong（香港大学）；Huawei Noah’s Ark Lab（华为诺亚方舟实验室）
The latent world model provides a promising way to learn policies in a compact latent space for tasks with high-dimensional observations, however, its generalization across diverse environments with unseen dynamics remains challenging. Although the recurrent structure utilized in current advances helps to capture local dynamics, modeling only state transitions without an explicit understanding of environmental context limits the generalization ability of the dynamics model. To address this issue, we propose a Prototypical Context-Aware Dynamics (ProtoCAD) model, which captures the local dynami…
Sparebank 1 SR-Bank；Center for Artificial Intelligence Research（人工智能研究中心）；University of Agder（阿格德大学）
The proliferation of artificial intelligence is increasingly dependent on model understanding. Understanding demands both an interpretation - a human reasoning about a model's behavior - and an explanation - a symbolic representation of the functioning of the model. Notwithstanding the imperative of transparency for safety, trust, and acceptance, the opacity of state-of-the-art reinforcement learning algorithms conceals the rudiments of their learned strategies. We have developed a policy regularization method that asserts the global intrinsic affinities of learned strategies. These affinities…
Reinforcement learning (RL) has demonstrated its ability to solve high dimensional tasks by leveraging non-linear function approximators. However, these successes are mostly achieved by 'black-box' policies in simulated domains. When deploying RL to the real world, several concerns regarding the use of a 'black-box' policy might be raised. In order to make the learned policies more transparent, we propose in this paper a policy iteration scheme that retains a complex function approximator for its internal value predictions but constrains the policy to have a concise, hierarchical, and human-re…
Reinforcement learning (RL) has been widely applied to game-playing and surpassed the best human-level performance in many domains, yet there are few use-cases in industrial or commercial settings. We introduce OR-Gym, an open-source library for developing reinforcement learning algorithms to address operations research problems. In this paper, we apply reinforcement learning to the knapsack, multi-dimensional bin packing, multi-echelon supply chain, and multi-period asset allocation model problems, as well as benchmark the RL solutions against MILP and heuristic models. These problems are use…
Spoken dialog systems have seen applications in many domains, including medical for automatic conversational diagnosis. State-of-the-art dialog managers are usually driven by deep reinforcement learning models, such as deep Q networks (DQNs), which learn by interacting with a simulator to explore the entire action space since real conversations are limited. However, the DQN-based automatic diagnosis models do not achieve satisfying performances when adapted to new, unseen diseases with only a few training samples. In this work, we propose the Prototypical Q Networks (ProtoQN) as the dialog man…
Most reinforcement learning methods operate on propositional representations of the world state. Such representations are often intractably large and generalize poorly. Using a deictic representation is believed to be a viable alternative: they promise generalization while allowing the use of existing reinforcement-learning methods. Yet, there are few experiments on learning with deictic representations reported in the literature. In this paper we explore the effectiveness of two forms of deictic representation and a naïve propositional representation in a simple blocks-world domain. We find, …
We present a generic and flexible Reinforcement Learning (RL) based meta-learning framework for the problem of few-shot learning. During training, it learns the best optimization algorithm to produce a learner (ranker/classifier, etc) by exploiting stable patterns in loss surfaces. Our method implicitly estimates the gradients of a scaled loss function while retaining the general properties intact for parameter updates. Besides providing improved performance on few-shot tasks, our framework could be easily extended to do network architecture search. We further propose a novel dual encoder, aff…
Department of Electrical-Engineering, Technion Institute of Technology, Israel（电气工程系，技术学院技术研究所，以色列）；Nvidia Research（Nvidia研究）；Technion (currently at Google Research)（技术学院（目前在谷歌研究））
Beijing Institute of Technology（北京理工大学）；University of the Sunshine Coast（阳光海岸大学）
This work explores how to learn robust and generalizable state representation from image-based observations with deep reinforcement learning methods. Addressing the computational complexity, stringent assumptions and representation collapse challenges in existing work of bisimulation metric, we devise Simple State Representation (SimSR) operator. SimSR enables us to design a stochastic approximation method that can practically learn the mapping functions (encoders) from observations to latent representation space. In addition to the theoretical analysis and comparison with the existing work, w…
Learning informative representations from image-based observations is of fundamental concern in deep Reinforcement Learning (RL). However, data-inefficiency remains a significant barrier to this objective. To overcome this obstacle, we propose to accelerate state representation learning by enforcing view-consistency on the dynamics. Firstly, we introduce a formalism of Multi-view Markov Decision Process (MMDP) that incorporates multiple views of the state. Following the structure of MMDP, our method, View-Consistent Dynamics (VCD), learns state representations by training a view-consistent dyn…
The options framework in reinforcement learning models the notion of a skill or a temporally extended sequence of actions. The discovery of a reusable set of skills has typically entailed building options, that navigate to bottleneck states. This work adopts a complementary approach, where we attempt to discover options that navigate to landmark states. These states are prototypical representatives of well-connected regions and can hence access the associated region with relative ease. In this work, we propose Successor Options, which leverages Successor Representations to build a model of the…
Knowledge Representation is important issue in reinforcement learning. In this paper, we bridge the gap between reinforcement learning and knowledge representation, by providing a rich knowledge representation framework, based on normal logic programs with answer set semantics, that is capable of solving model-free reinforcement learning problems for more complex do-mains and exploits the domain-specific knowledge. We prove the correctness of our approach. We show that the complexity of finding an offline and online policy for a model-free reinforcement learning problem in our approach is NP-c…
INESC-ID（里斯本高等理工学院信息技术与系统研究所）；Instituto Superior Técnico, Universidade de Lisboa（里斯本大学高等理工学院）；KTH Royal Institute of Technology（瑞典皇家理工学院）
This work addresses the problem of sensing the world: how to learn a multimodal representation of a reinforcement learning agent's environment that allows the execution of tasks under incomplete perceptual conditions. To address such problem, we argue for hierarchy in the design of representation models and contribute with a novel multimodal representation model, MUSE. The proposed model learns hierarchical representations: low-level modality-specific representations, encoded from raw observation data, and a high-level multimodal representation, encoding joint-modality information to allow rob…
Learning a good representation is an essential component for deep reinforcement learning (RL). Representation learning is especially important in multitask and partially observable settings where building a representation of the unknown environment is crucial to solve the tasks. Here we introduce Prediction of Bootstrap Latents (PBL), a simple and flexible self-supervised representation learning algorithm for multitask deep RL. PBL builds on multistep predictive representations of future observations, and focuses on capturing structured information about environment dynamics. Specifically, PBL…
Locally Constrained Representations in Reinforcement Learning
McGill University（麦吉尔大学）；Mila-Quebec AI Institute（米拉-魁北克人工智能研究所）；University of Massachusetts, Amherst（马萨诸塞大学阿默斯特分校）；Dell Technologies, Research Office（戴尔科技公司研究办公室）；École de technologie supérieure（高等技术学院）；CIFAR AI Chair（加拿大高级研究所人工智能讲席）
National University of Singapore（新加坡国立大学）；ByteDance（字节跳动）
In Reinforcement Learning (RL), Laplacian Representation (LapRep) is a task-agnostic state representation that encodes the geometry of the environment. A desirable property of LapRep stated in prior works is that the Euclidean distance in the LapRep space roughly reflects the reachability between states, which motivates the usage of this distance for reward shaping. However, we find that LapRep does not necessarily have this property in general: two states having small distance under LapRep can actually be far away in the environment. Such mismatch would impede the learning process in reward s…
Reinforcement Learning (RL) can be considered as a sequence modeling task: given a sequence of past state-action-reward experiences, an agent predicts a sequence of next actions. In this work, we propose State-Action-Reward Transformer (StARformer) for visual RL, which explicitly models short-term state-action-reward representations (StAR-representations), essentially introducing a Markovian-like inductive bias to improve long-term modeling. Our approach first extracts StAR-representations by self-attending image state patches, action, and reward tokens within a short temporal window. These ar…
Beijing Institute of Technology（北京理工大学）；MILA（蒙特利尔学习算法研究所）；Microsoft Research Montreal（微软研究院蒙特利尔）
Offline reinforcement learning (RL) struggles in environments with rich and noisy inputs, where the agent only has access to a fixed dataset without environment interactions. Past works have proposed common workarounds based on the pre-training of state representations, followed by policy training. In this work, we introduce a simple, yet effective approach for learning state representations. Our method, Behavior Prior Representation (BPR), learns state representations with an easy-to-integrate objective based on behavior cloning of the dataset: we first learn a state representation by mimicki…
Here, we report a case study implementation of reinforcement learning (RL) to automate operations in the scanning transmission electron microscopy (STEM) workflow. To do so, we design a virtual, prototypical RL environment to test and develop a network to autonomously align the electron beam without prior knowledge. Using this simulator, we evaluate the impact of environment design and algorithm hyperparameters on alignment accuracy and learning convergence, showing robust convergence across a wide hyperparameter space. Additionally, we deploy a successful model on the microscope to validate t…
Center for Data Science（数据科学中心）；New York University（纽约大学）
We explore the benefits of augmenting state-of-the-art model-free deep reinforcement algorithms with simple object representations. Following the Frostbite challenge posited by Lake et al. (2017), we identify object representations as a critical cognitive capacity lacking from current reinforcement learning agents. We discover that providing the Rainbow model (Hessel et al.,2018) with simple, feature-engineered object representations substantially boosts its performance on the Frostbite game from Atari 2600. We then analyze the relative contributions of the representations of different types o…
Reinforcement Learning (RL) of contact-rich manipulation tasks has yielded impressive results in recent years. While many studies in RL focus on varying the observation space or reward model, few efforts focused on the choice of action space (e.g. joint or end-effector space, position, velocity, etc.). However, studies in robot motion control indicate that choosing an action space that conforms to the characteristics of the task can simplify exploration and improve robustness to disturbances. This paper studies the effect of different action spaces in deep RL and advocates for Variable Impedan…
A variety of representation learning approaches have been investigated for reinforcement learning; much less attention, however, has been given to investigating the utility of sparse coding. Outside of reinforcement learning, sparse coding representations have been widely used, with non-convex objectives that result in discriminative representations. In this work, we develop a supervised sparse coding objective for policy evaluation. Despite the non-convexity of this objective, we prove that all local minima are global minima, making the approach amenable to simple optimization strategies. We …
University of Alberta（阿尔伯塔大学）；Indiana University Bloomington（印第安纳大学伯明顿分校）
We investigate sparse representations for control in reinforcement learning. While these representations are widely used in computer vision, their prevalence in reinforcement learning is limited to sparse coding where extracting representations for new data can be computationally intensive. Here, we begin by demonstrating that learning a control policy incrementally with a representation from a standard neural network fails in classic control domains, whereas learning with a representation obtained from a neural network that has sparsity properties enforced is effective. We provide evidence th…
This paper presents a preliminary study comparing different observation and action space representations for Deep Reinforcement Learning (DRL) in the context of Real-time Strategy (RTS) games. Specifically, we compare two representations: (1) a global representation where the observation represents the whole game state, and the RL agent needs to choose which unit to issue actions to, and which actions to execute; and (2) a local representation where the observation is represented from the point of view of an individual unit, and the RL agent picks actions for each unit independently. We evalua…
Poor sample efficiency is a major limitation of deep reinforcement learning in many domains. This work presents an attention-based method to project neural network inputs into an efficient representation space that is invariant under changes to input ordering. We show that our proposed representation results in an input space that is a factor of $m!$ smaller for inputs of $m$ objects. We also show that our method is able to represent inputs over variable numbers of objects. Our experiments demonstrate improvements in sample efficiency for policy gradient methods on a variety of tasks. We show …
University of Washington（华盛顿大学）；Microsoft（微软）；Criteo Research（Criteo研究院）
Deep reinforcement learning is quickly changing the field of artificial intelligence. These models are able to capture a high level understanding of their environment, enabling them to learn difficult dynamic tasks in a variety of domains. In the database field, query optimization remains a difficult problem. Our goal in this work is to explore the capabilities of deep reinforcement learning in the context of query optimization. At each state, we build queries incrementally and encode properties of subqueries through a learned representation. The challenge here lies in the formation of the sta…
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…
The prototypical approach to reinforcement learning involves training policies tailored to a particular agent from scratch for every new morphology. Recent work aims to eliminate the re-training of policies by investigating whether a morphology-agnostic policy, trained on a diverse set of agents with similar task objectives, can be transferred to new agents with unseen morphologies without re-training. This is a challenging problem that required previous approaches to use hand-designed descriptions of the new agent's morphology. Instead of hand-designing this description, we propose a data-dri…
University of Toronto（多伦多大学）；Barclays Capital（巴克莱资本）
We present a reinforcement learning (RL) approach for robust optimisation of risk-aware performance criteria. To allow agents to express a wide variety of risk-reward profiles, we assess the value of a policy using rank dependent expected utility (RDEU). RDEU allows the agent to seek gains, while simultaneously protecting themselves against downside risk. To robustify optimal policies against model uncertainty, we assess a policy not by its distribution, but rather, by the worst possible distribution that lies within a Wasserstein ball around it. Thus, our problem formulation may be viewed as …
McGill University（麦吉尔大学）；Mila；DeepMind；Facebook AI Research（Facebook人工智能研究院）
In the quest for efficient and robust reinforcement learning methods, both model-free and model-based approaches offer advantages. In this paper we propose a new way of explicitly bridging both approaches via a shared low-dimensional learned encoding of the environment, meant to capture summarizing abstractions. We show that the modularity brought by this approach leads to good generalization while being computationally efficient, with planning happening in a smaller latent state space. In addition, this approach recovers a sufficient low-dimensional representation of the environment, which op…
Offline reinforcement learning, which seeks to utilize offline/historical data to optimize sequential decision-making strategies, has gained surging prominence in recent studies. Due to the advantage that appropriate function approximators can help mitigate the sample complexity burden in modern reinforcement learning problems, existing endeavors usually enforce powerful function representation models (e.g. neural networks) to learn the optimal policies. However, a precise understanding of the statistical limits with function representations, remains elusive, even when such a representation is…
One question central to Reinforcement Learning is how to learn a feature representation that supports algorithm scaling and re-use of learned information from different tasks. Successor Features approach this problem by learning a feature representation that satisfies a temporal constraint. We present an implementation of an approach that decouples the feature representation from the reward function, making it suitable for transferring knowledge between domains. We then assess the advantages and limitations of using Successor Features for transfer.
University of Oxford（牛津大学）；Google Brain（谷歌大脑）；McGill University（麦吉尔大学）
In reinforcement learning, state representations are used to tractably deal with large problem spaces. State representations serve both to approximate the value function with few parameters, but also to generalize to newly encountered states. Their features may be learned implicitly (as part of a neural network) or explicitly (for example, the successor representation of \citet{dayan1993improving}). While the approximation properties of representations are reasonably well-understood, a precise characterization of how and when these representations generalize is lacking. In this work, we addres…
The objective of transfer reinforcement learning is to generalize from a set of previous tasks to unseen new tasks. In this work, we focus on the transfer scenario where the dynamics among tasks are the same, but their goals differ. Although general value function (Sutton et al., 2011) has been shown to be useful for knowledge transfer, learning a universal value function can be challenging in practice. To attack this, we propose (1) to use universal successor representations (USR) to represent the transferable knowledge and (2) a USR approximator (USRA) that can be trained by interacting with…
While deep reinforcement learning has proven to be successful in solving control tasks, the "black-box" nature of an agent has received increasing concerns. We propose a prototype-based post-hoc policy explainer, ProtoX, that explains a blackbox agent by prototyping the agent's behaviors into scenarios, each represented by a prototypical state. When learning prototypes, ProtoX considers both visual similarity and scenario similarity. The latter is unique to the reinforcement learning context, since it explains why the same action is taken in visually different states. To teach ProtoX about vis…
Institute for Advanced Study（高等研究院）；University of Washington, Seattle（西雅图华盛顿大学）；Carnegie Mellon University（卡内基梅隆大学）；University of California, Los Angles（加利福尼亚大学洛杉矶分校）
Modern deep learning methods provide effective means to learn good representations. However, is a good representation itself sufficient for sample efficient reinforcement learning? This question has largely been studied only with respect to (worst-case) approximation error, in the more classical approximate dynamic programming literature. With regards to the statistical viewpoint, this question is largely unexplored, and the extant body of literature mainly focuses on conditions which permit sample efficient reinforcement learning with little understanding of what are necessary conditions for …
National University of Singapore（新加坡国立大学）；City University of Hong Kong（香港城市大学）；ByteDance AI lab（字节跳动人工智能实验室）
The Laplacian representation recently gains increasing attention for reinforcement learning as it provides succinct and informative representation for states, by taking the eigenvectors of the Laplacian matrix of the state-transition graph as state embeddings. Such representation captures the geometry of the underlying state space and is beneficial to RL tasks such as option discovery and reward shaping. To approximate the Laplacian representation in large (or even continuous) state spaces, recent works propose to minimize a spectral graph drawing objective, which however has infinitely many g…
We study the problem of representation learning in goal-conditioned hierarchical reinforcement learning. In such hierarchical structures, a higher-level controller solves tasks by iteratively communicating goals which a lower-level policy is trained to reach. Accordingly, the choice of representation -- the mapping of observation space to goal space -- is crucial. To study this problem, we develop a notion of sub-optimality of a representation, defined in terms of expected reward of the optimal hierarchical policy using this representation. We derive expressions which bound the sub-optimality …
This paper presents a novel state representation for reward-free Markov decision processes. The idea is to learn, in a self-supervised manner, an embedding space where distances between pairs of embedded states correspond to the minimum number of actions needed to transition between them. Compared to previous methods, our approach does not require any domain knowledge, learning from offline and unlabeled data. We show how this representation can be leveraged to learn goal-conditioned policies, providing a notion of similarity between states and goals and a useful heuristic distance to guide pl…
We consider a reinforcement learning setting introduced in (Maillard et al., NIPS 2011) where the learner does not have explicit access to the states of the underlying Markov decision process (MDP). Instead, she has access to several models that map histories of past interactions to states. Here we improve over known regret bounds in this setting, and more importantly generalize to the case where the models given to the learner do not contain a true model resulting in an MDP representation but only approximations of it. We also give improved error bounds for state aggregation.
Deep Reinforcement learning is a branch of unsupervised learning in which an agent learns to act based on environment state in order to maximize its total reward. Deep reinforcement learning provides good opportunity to model the complexity of portfolio choice in high-dimensional and data-driven environment by leveraging the powerful representation of deep neural networks. In this paper, we build a portfolio management system using direct deep reinforcement learning to make optimal portfolio choice periodically among S\&P500 underlying stocks by learning a good factor representation (as input)…
UCL（伦敦大学学院）；Huawei Noah’s Ark Lab（华为诺亚方舟实验室）；College of Intelligence and Computing, Tianjin University（天津大学智能与计算学部）
We propose learning via retracing, a novel self-supervised approach for learning the state representation (and the associated dynamics model) for reinforcement learning tasks. In addition to the predictive (reconstruction) supervision in the forward direction, we propose to include "retraced" transitions for representation / model learning, by enforcing the cycle-consistency constraint between the original and retraced states, hence improve upon the sample efficiency of learning. Moreover, learning via retracing explicitly propagates information about future transitions backward for inferring …
Institute of Computational and Mathematical Engineering（计算与数学工程研究所）
Interactive adaptive systems powered by Reinforcement Learning (RL) have many potential applications, such as intelligent tutoring systems. In such systems there is typically an external human system designer that is creating, monitoring and modifying the interactive adaptive system, trying to improve its performance on the target outcomes. In this paper we focus on algorithmic foundation of how to help the system designer choose the set of sensors or features to define the observation space used by reinforcement learning agent. We present an algorithm, value driven representation (VDR), that …
The problem of selecting the right state-representation in a reinforcement learning problem is considered. Several models (functions mapping past observations to a finite set) of the observations are given, and it is known that for at least one of these models the resulting state dynamics are indeed Markovian. Without knowing neither which of the models is the correct one, nor what are the probabilistic characteristics of the resulting MDP, it is required to obtain as much reward as the optimal policy for the correct model (or for the best of the correct models, if there are several). We propo…
Argonne National Laboratory（阿贡国家实验室）；University of Southern California（南加州大学）
Predictive materials synthesis is the primary bottleneck in realizing new functional and quantum materials. Strategies for synthesis of promising materials are currently identified by time-consuming trial and error approaches and there are no known predictive schemes to design synthesis parameters for new materials. We use reinforcement learning to predict optimal synthesis schedules, i.e. a time-sequence of reaction conditions like temperatures and reactant concentrations, for the synthesis of a prototypical quantum material, semiconducting monolayer MoS$_{2}$, using chemical vapor deposition…
MINES ParisTech（巴黎高科国立高等矿业学校）；PSL Research University（PSL研究大学）；Centre de mise en forme des matériaux (CEMEF)（材料成型中心（CEMEF））；CNRS UMR 7635（法国国家科学研究中心7635联合研究单位）
This research gauges the ability of deep reinforcement learning (DRL) techniques to assist the optimization and control of fluid mechanical systems. It combines a novel, "degenerate" version of the proximal policy optimization (PPO) algorithm, that trains a neural network in optimizing the system only once per learning episode, and an in-house stabilized finite elements environment implementing the variational multiscale (VMS) method, that computes the numerical reward fed to the neural network. Three prototypical examples of separated flows in two dimensions are used as testbed for developing…
University of Science and Technology of China（中国科学技术大学）；SUNY Binghamton（纽约州立大学宾汉姆顿分校）；UT Austin（德克萨斯大学奥斯汀分校）
Reinforcement learning (RL) agents aim at learning by interacting with an environment, and are not designed for representing or reasoning with declarative knowledge. Knowledge representation and reasoning (KRR) paradigms are strong in declarative KRR tasks, but are ill-equipped to learn from such experiences. In this work, we integrate logical-probabilistic KRR with model-based RL, enabling agents to simultaneously reason with declarative knowledge and learn from interaction experiences. The knowledge from humans and RL is unified and used for dynamically computing task-specific planning model…
