# 61 Interesting Paper from NeurIPS 2019

[reinforcement-learning

deep-learning

]
# Reinforcement Learning and Planning – Reinforcement Learning

## *A Geometric Perspective on Optimal Representations for Reinforcement Learning

Marc Bellemare · Will Dabney · Robert Dadashi · Adrien Ali Taiga · Pablo Samuel Castro · Nicolas Le Roux · Dale Schuurmans · Tor Lattimore · Clare Lyle We propose a new perspective on representation learning in reinforcement learning based on geometric properties of the space of value functions. From there, we provide formal evidence regarding the usefulness of value functions as auxiliary tasks in reinforcement learning. Our formulation considers adapting the representation to minimize the (linear) approximation of the value function of all stationary policies for a given environment. We show that this optimization reduces to making accurate predictions regarding a special class of value functions which we call adversarial value functions (AVFs). We demonstrate that using value functions as auxiliary tasks corresponds to an expected-error relaxation of our formulation, with AVFs a natural candidate, and identify a close relationship with proto-value functions (Mahadevan, 2005). We highlight characteristics of AVFs and their usefulness as auxiliary tasks in a series of experiments on the four-room domain.

## Multi-View Reinforcement Learning

Minne Li · Lisheng Wu · Jun WANG · Haitham Bou Ammar This paper is concerned with multi-view reinforcement learning (MVRL), which allows for decision making when agents share common dynamics but adhere to different observation models. We define the MVRL framework by extending partially observable Markov decision processes (POMDPs) to support more than one observation model and propose two solution methods through observation augmentation and cross-view policy transfer. We empirically evaluate our method and demonstrate its effectiveness in a variety of environments. Specifically, we show reductions in sample complexities and computational time for acquiring policies that handle multi-view environments.

## Sample-Efficient Deep Reinforcement Learning via Episodic Backward Update

Su Young Lee · Choi Sungik · Sae-Young Chung We propose Episodic Backward Update (EBU) – a novel deep reinforcement learning algorithm with a direct value propagation. In contrast to the conventional use of the experience replay with uniform random sampling, our agent samples a whole episode and successively propagates the value of a state to its previous states. Our computationally efficient recursive algorithm allows sparse and delayed rewards to propagate directly through all transitions of the sampled episode. We theoretically prove the convergence of the EBU method and experimentally demonstrate its performance in both deterministic and stochastic environments. Especially in 49 games of Atari 2600 domain, EBU achieves the same mean and median human normalized performance of DQN by using only 5% and 10% of samples, respectively.

## Information-Theoretic Confidence Bounds for Reinforcement Learning

Xiuyuan Lu · Benjamin Van Roy We integrate information-theoretic concepts into the design and analysis of optimistic algorithms and Thompson sampling. By making a connection between information-theoretic quantities and confidence bounds, we obtain results that relate the per-period performance of the agent with its information gain about the environment, thus explicitly characterizing the exploration-exploitation tradeoff. The resulting cumulative regret bound depends on the agent’s uncertainty over the environment and quantifies the value of prior information. We show applicability of this approach to several environments, including linear bandits, tabular MDPs, and factored MDPs. These examples demonstrate the potential of a general information-theoretic approach for the design and analysis of reinforcement learning algorithms.

## Regret Minimization for Reinforcement Learning by Evaluating the Optimal Bias Function

Zihan Zhang · Xiangyang Ji We present an algorithm based on the \emph{Optimism in the Face of Uncertainty} (OFU) principle which is able to learn Reinforcement Learning (RL) modeled by Markov decision process (MDP) with finite state-action space efficiently. By evaluating the state-pair difference of the optimal bias function h∗, the proposed algorithm achieves a regret bound of ~O(√SATH)\footnote{The symbol ~O means O with log factors ignored. } for MDP with S states and A actions, in the case that an upper bound H on the span of h∗, i.e., sp(h∗) is known. This result outperforms the best previous regret bounds ~O(HS√AT)\cite{bartlett2009regal} by a factor of √SH. Furthermore, this regret bound matches the lower bound of Ω(√SATH)\cite{jaksch2010near} up to a logarithmic factor. As a consequence, we show that there is a near optimal regret bound of ~O(√DSAT) for MDPs with finite diameter D compared to the lower bound of Ω(√DSAT)\cite{jaksch2010near}.

## Real-Time Reinforcement Learning

Simon Ramstedt · Chris Pal Markov Decision Processes (MDPs), the mathematical framework underlying most algorithms in Reinforcement Learning (RL), are often used in a way that wrongfully assumes that the state of an agent’s environment does not change during action selection. As RL systems based on MDPs begin to find application in real-world safety critical situations, this mismatch between the assumptions underlying classical MDPs and the reality of real-time computation may lead to undesirable outcomes. In this paper, we introduce a new framework, in which states and actions evolve simultaneously and show how it is related to the classical MDP formulation. We analyze existing algorithms under the new real-time formulation and show why they are suboptimal when used in real-time. We then use those insights to create a new algorithm Real-Time Actor Critic (RTAC) that outperforms the existing state-of-the-art continuous control algorithm Soft Actor Critic both in real-time and non-real-time settings.

## Convergent Policy Optimization for Safe Reinforcement Learning

Ming Yu · Zhuoran Yang · Mladen Kolar · Zhaoran Wang We study the safe reinforcement learning problem with nonlinear function approximation, where policy optimization is formulated as a constrained optimization problem with both the objective and the constraint being nonconvex functions. For such a problem, we construct a sequence of surrogate convex constrained optimization problems by replacing the nonconvex functions locally with convex quadratic functions obtained from policy gradient estimators. We prove that the solutions to these surrogate problems converge to a stationary point of the original nonconvex problem. Furthermore, to extend our theoretical results, we apply our algorithm to examples of optimal control and multi-agent reinforcement learning with safety constraints.

## Intrinsically Efficient, Stable, and Bounded Off-Policy Evaluation for Reinforcement Learning

Nathan Kallus · Masatoshi Uehara Off-policy evaluation (OPE) in both contextual bandits and reinforcement learning allows one to evaluate novel decision policies without needing to conduct exploration, which is often costly or otherwise infeasible. The problem’s importance has attracted many proposed solutions, including importance sampling (IS), self-normalized IS (SNIS), and doubly robust (DR) estimates. DR and its variants ensure semiparametric local efficiency if Q-functions are well-specified, but if they are not they can be worse than both IS and SNIS. It also does not enjoy SNIS’s inherent stability and boundedness. We propose new estimators for OPE based on empirical likelihood that are always more efficient than IS, SNIS, and DR and satisfy the same stability and boundedness properties as SNIS. On the way, we categorize various properties and classify existing estimators by them. Besides the theoretical guarantees, empirical studies suggest the new estimators provide advantages.

## Propagating Uncertainty in Reinforcement Learning via Wasserstein Barycenters

Alberto Maria Metelli · Amarildo Likmeta · Marcello Restelli How does the uncertainty of the value function propagate when performing temporal difference learning? In this paper, we address this question by proposing a Bayesian framework in which we employ approximate posterior distributions to model the uncertainty of the value function and Wasserstein barycenters to propagate it across state-action pairs. Leveraging on these tools, we present an algorithm, Wasserstein Q-Learning (WQL), starting in the tabular case and then, we show how it can be extended to deal with continuous domains. Furthermore, we prove that, under mild assumptions, a slight variation of WQL enjoys desirable theoretical properties in the tabular setting. Finally, we present an experimental campaign to show the effectiveness of WQL on finite problems, compared to several RL algorithms, some of which are specifically designed for exploration, along with some preliminary results on Atari games.

## Finite-Time Performance Bounds and Adaptive Learning Rate Selection for Two Time-Scale Reinforcement Learning

Harsh Gupta · R. Srikant · Lei Ying We study two time-scale linear stochastic approximation algorithms, which can be used to model well-known reinforcement learning algorithms such as GTD, GTD2, and TDC. We present finite-time performance bounds for the case where the learning rate is fixed. The key idea in obtaining these bounds is to use a Lyapunov function motivated by singular perturbation theory for linear differential equations. We use the bound to design an adaptive learning rate scheme which significantly improves the convergence rate over the known optimal polynomial decay rule in our experiments, and can be used to potentially improve the performance of any other schedule where the learning rate is changed at pre-determined time instants.

## Interval timing in deep reinforcement learning agents

Ben Deverett · Ryan Faulkner · Meire Fortunato · Gregory Wayne · Joel Leibo The measurement of time is central to intelligent behavior. We know that both animals and artificial agents can successfully use temporal dependencies to select actions. In artificial agents, little work has directly addressed (1) which architectural components are necessary for successful development of this ability, (2) how this timing ability comes to be represented in the units and actions of the agent, and (3) whether the resulting behavior of the system converges on solutions similar to those of biology. Here we studied interval timing abilities in deep reinforcement learning agents trained end-to-end on an interval reproduction paradigm inspired by experimental literature on mechanisms of timing. We characterize the strategies developed by recurrent and feedforward agents, which both succeed at temporal reproduction using distinct mechanisms, some of which bear specific and intriguing similarities to biological systems. These findings advance our understanding of how agents come to represent time, and they highlight the value of experimentally inspired approaches to characterizing agent abilities.

## Non-Stationary Markov Decision Processes, a Worst-Case Approach using Model-Based Reinforcement Learning

Erwan Lecarpentier · Emmanuel Rachelson This work tackles the problem of robust zero-shot planning in non-stationary stochastic environments. We study Markov Decision Processes (MDPs) evolving over time and consider Model-Based Reinforcement Learning algorithms in this setting. We make two hypotheses: 1) the environment evolves continuously with a bounded evolution rate; 2) a current model is known at each decision epoch but not its evolution. Our contribution can be presented in four points. 1) we define a specific class of MDPs that we call Non-Stationary MDPs (NSMDPs). We introduce the notion of regular evolution by making an hypothesis of Lipschitz-Continuity on the transition and reward functions w.r.t. time; 2) we consider a planning agent using the current model of the environment but unaware of its future evolution. This leads us to consider a worst-case method where the environment is seen as an adversarial agent; 3) following this approach, we propose the Risk-Averse Tree-Search (RATS) algorithm, a zero-shot Model-Based method similar to Minimax search; 4) we illustrate the benefits brought by RATS empirically and compare its performance with reference Model-Based algorithms.

## Budgeted Reinforcement Learning in Continuous State Space

Nicolas Carrara · Edouard Leurent · Romain Laroche · Tanguy Urvoy · Odalric-Ambrym Maillard · Olivier Pietquin A Budgeted Markov Decision Process (BMDP) is an extension of a Markov Decision Process to critical applications requiring safety constraints. It relies on a notion of risk implemented in the shape of an upper bound on a constrains violation signal that – importantly – can be modified in real-time. So far, BMDPs could only be solved in the case of finite state spaces with known dynamics. This work extends the state-of-the-art to continuous spaces environments and unknown dynamics. We show that the solution to a BMDP is the fixed point of a novel Budgeted Bellman Optimality operator. This observation allows us to introduce natural extensions of Deep Reinforcement Learning algorithms to address large-scale BMDPs. We validate our approach on two simulated applications: spoken dialogue and autonomous driving.

## Regularized Anderson Acceleration for Off-Policy Deep Reinforcement Learning

Wenjie Shi · Shiji Song · Hui Wu · Ya-Chu Hsu · Cheng Wu · Gao Huang Model-free deep reinforcement learning (RL) algorithms have been widely used for a range of complex control tasks. However, slow convergence and sample inefficiency remain challenging problems in RL, especially when handling continuous and high-dimensional state spaces. To tackle this problem, we propose a general acceleration method for model-free, off-policy deep RL algorithms by drawing the idea underlying regularized Anderson acceleration (RAA), which is an effective approach to accelerating the solving of fixed point problems with perturbations. Specifically, we first explain how policy iteration can be applied directly with Anderson acceleration. Then we extend RAA to the case of deep RL by introducing a regularization term to control the impact of perturbation induced by function approximation errors. We further propose two strategies, i.e., progressive update and adaptive restart, to enhance the performance. The effectiveness of our method is evaluated on a variety of benchmark tasks, including Atari 2600 and MuJoCo. Experimental results show that our approach substantially improves both the learning speed and final performance of state-of-the-art deep RL algorithms.

## Tight Regret Bounds for Model-Based Reinforcement Learning with Greedy Policies

Yonathan Efroni · Nadav Merlis · Mohammad Ghavamzadeh · Shie Mannor State-of-the-art efficient model-based Reinforcement Learning (RL) algorithms typically act by iteratively solving empirical models, i.e., by performing full-planning on Markov Decision Processes (MDPs) built by the gathered experience. In this paper, we focus on model-based RL in the finite-state finite-horizon MDP setting and establish that exploring with greedy policies – act by 1-step planning – can achieve tight minimax performance in terms of regret, O(\sqrt{HSAT}). Thus, full-planning in model-based RL can be avoided altogether without any performance degradation, and, by doing so, the computational complexity decreases by a factor of S. The results are based on a novel analysis of real-time dynamic programming, then extended to model-based RL. Specifically, we generalize existing algorithms that perform full-planning to such that act by 1-step planning. For these generalizations, we prove regret bounds with the same rate as their full-planning counterparts.

## Regret Bounds for Learning State Representations in Reinforcement Learning

Ronald Ortner · Matteo Pirotta · Alessandro Lazaric · Ronan Fruit · Odalric-Ambrym Maillard We consider the problem of online reinforcement learning when several state representations (mapping histories to a discrete state space) are available to the learning agent. At least one of these representations is assumed to induce a Markov decision process (MDP), and the performance of the agent is measured in terms of cumulative regret against the optimal policy giving the highest average reward in this MDP representation. We propose an algorithm (UCB-MS) with O(sqrt(T)) regret in any communicating Markov decision process. The regret bound shows that UCB-MS automatically adapts to the Markov model. This improves over the currently known best results in the literature that gave regret bounds of order O(T^(2/3)).

## Reinforcement Learning with Convex Constraints

Sobhan Miryoosefi · Kianté Brantley · Hal Daume III · Miro Dudik · Robert Schapire In standard reinforcement learning (RL), a learning agent seeks to optimize the overall reward. However, many key aspects of a desired behavior are more naturally expressed as constraints. For instance, the designer may want to limit the use of unsafe actions, increase the diversity of trajectories to enable exploration, or approximate expert trajectories when rewards are sparse. In this paper, we propose an algorithmic scheme that can handle a wide class of constraints in RL tasks: specifically, any constraints that require expected values of some vector measurements (such as the use of an action) to lie in a convex set. This captures previously studied constraints (such as safety and proximity to an expert), but also enables new classes of constraints (such as diversity). Our approach comes with rigorous theoretical guarantees and only relies on the ability to approximately solve standard RL tasks. As a result, it can be easily adapted to work with any model-free or model-based RL. In our experiments, we show that it matches previous algorithms that enforce safety via constraints, but can also enforce new properties that these algorithms do not incorporate, such as diversity.

## Correlation Priors for Reinforcement Learning

Bastian Alt · Adrian Šošić · Heinz Koeppl Many decision-making problems naturally exhibit pronounced structures inherited from the characteristics of the underlying environment. In a Markov decision process model, for example, two distinct states can have inherently related semantics or encode resembling physical state configurations. This often implies locally correlated transition dynamics among the states. In order to complete a certain task in such environments, the operating agent usually needs to execute a series of temporally and spatially correlated actions. Though there exists a variety of approaches to capture these correlations in continuous state-action domains, a principled solution for discrete environments is missing. In this work, we present a Bayesian learning framework based on Pólya-Gamma augmentation that enables an analogous reasoning in such cases. We demonstrate the framework on a number of common decision-making related problems, such as imitation learning, subgoal extraction, system identification and Bayesian reinforcement learning. By explicitly modeling the underlying correlation structures of these problems, the proposed approach yields superior predictive performance compared to correlation-agnostic models, even when trained on data sets that are an order of magnitude smaller in size.

## Policy Poisoning in Batch Reinforcement Learning and Control

Yuzhe Ma · Xuezhou Zhang · Wen Sun · Jerry Zhu We study a security threat to batch reinforcement learning and control where the attacker aims to poison the learned policy. The victim is a reinforcement learner / controller which first estimates the dynamics and the rewards from a batch data set, and then solves for the optimal policy with respect to the estimates. The attacker can modify the data set slightly before learning happens, and wants to force the learner into learning a target policy chosen by the attacker. We present a unified framework for solving batch policy poisoning attacks, and instantiate the attack on two standard victims: tabular certainty equivalence learner in reinforcement learning and linear quadratic regulator in control. We show that both instantiation result in a convex optimization problem on which global optimality is guaranteed, and provide analysis on attack feasibility and attack cost. Experiments show the effectiveness of policy poisoning attacks.

# Learning Tricks

## *When to use parametric models in reinforcement learning

Hado van Hasselt · Matteo Hessel · John Aslanides We examine the question of when and how parametric models are most useful in reinforcement learning. In particular, we look at commonalities and differences between parametric models and experience replay. Replay-based learning algorithms share important traits with model-based approaches, including the ability to plan: to use more computation without additional data to improve predictions and behaviour. We discuss when to expect benefits from either approach, and interpret prior work in this context. We hypothesise that, under suitable conditions, replay-based algorithms should be competitive to or better than model-based algorithms if the model is used only to generate fictional transitions from observed states for an update rule that is otherwise model-free. We validated this hypothesis on Atari 2600 video games. The replay-based algorithm attained state-of-the-art data efficiency, improving over prior results with parametric models. Additionally, we discuss different ways to use models. We show that it can be better to plan backward than to plan forward when using models to perform credit assignment (e.g., to directly learn a value or policy), even though the latter seems more common. Finally, we argue and demonstrate that it can be beneficial to plan forward for immediate behaviour, rather than for credit assignment.

## The Option Keyboard: Combining Skills in Reinforcement Learning

Daniel Toyama · Shaobo Hou · Gheorghe Comanici · Andre Barreto · Doina Precup · Shibl Mourad · Eser Aygün · Philippe Hamel Our paper introduces a modular RL algorithm that provides a temporally extended interface for RL agents, akin to a piano keyboard: the agent chooses among a large selection of “chords” that correspond to linear combinations of “keys” executed over an extended number of environment steps. The added level of abstraction is obtained by pre-training a set of skills corresponding to a finite set of chords and generalized policy evaluation and improvement to synthesize any other chord on-the-fly. We would like to demonstrate the flexibility of the proposed interface by allowing audience members to perform complex RL tasks through the use of a combination of a small set of skills corresponding to intuitive short term objectives. MIDI musical keyboards will be used to control virtual physical bodies through the original action space (i.e. control body joints), as well as abstract action interfaces. The latter concretely illustrates the ability of linearly combining a finite set of skills, akin to playing chords using a small number of keys.

## Robust exploration in linear quadratic reinforcement learning

Jack Umenberger · Mina Ferizbegovic · Thomas Schön · Håkan Hjalmarsson Learning to make decisions in an uncertain and dynamic environment is a task of fundamental performance in a number of domains. This paper concerns the problem of learning control policies for an unknown linear dynamical system so as to minimize a quadratic cost function. We present a method, based on convex optimization, that accomplishes this task ‘robustly’, i.e., the worst-case cost, accounting for system uncertainty given the observed data, is minimized. The method balances exploitation and exploration, exciting the system in such a way so as to reduce uncertainty in the model parameters to which the worst-case cost is most sensitive. Numerical simulations and application to a hardware-in-the-loop servo-mechanism are used to demonstrate the approach, with appreciable performance and robustness gains over alternative methods observed in both.

# Framework

## A Regularized Approach to Sparse Optimal Policy in Reinforcement Learning

Wenhao Yang · Xiang Li · Zhihua Zhang We propose and study a general framework for regularized Markov decision processes (MDPs) where the goal is to find an optimal policy that maximizes the expected discounted total reward plus a policy regularization term. The extant entropy-regularized MDPs can be cast into our framework. Moreover, under our framework, many regularization terms can bring multi-modality and sparsity, which are potentially useful in reinforcement learning. In particular, we present sufficient and necessary conditions that induce a sparse optimal policy. We also conduct a full mathematical analysis of the proposed regularized MDPs, including the optimality condition, performance error, and sparseness control. We provide a generic method to devise regularization forms and propose off-policy actor critic algorithms in complex environment settings. We empirically analyze the numerical properties of optimal policies and compare the performance of different sparse regularization forms in discrete and continuous environments.

## VIREL: A Variational Inference Framework for Reinforcement Learning

Matthew Fellows · Anuj Mahajan · Tim G. J. Rudner · Shimon Whiteson Applying probabilistic models to reinforcement learning (RL) enables the uses of powerful optimisation tools such as variational inference in RL. However, existing inference frameworks and their algorithms pose significant challenges for learning optimal policies, e.g., the lack of mode capturing behaviour in pseudo-likelihood methods, difficulties learning deterministic policies in maximum entropy RL based approaches, and a lack of analysis when function approximators are used. We propose VIREL, a theoretically grounded probabilistic inference framework for RL that utilises a parametrised action-value function to summarise future dynamics of the underlying MDP, generalising existing approaches. VIREL also benefits from a mode-seeking form of KL divergence, the ability to learn deterministic optimal polices naturally from inference, and the ability to optimise value functions and policies in separate, iterative steps. In applying variational expectation-maximisation to VIREL, we thus show that the actor-critic algorithm can be reduced to expectation-maximisation, with policy improvement equivalent to an E-step and policy evaluation to an M-step. We then derive a family of actor-critic methods fromVIREL, including a scheme for adaptive exploration. Finally, we demonstrate that actor-critic algorithms from this family outperform state-of-the-art methods based on soft value functions in several domains.

## Gossip-based Actor-Learner Architectures for Deep Reinforcement Learning

Mahmoud (“Mido”) Assran · Joshua Romoff · Nicolas Ballas · Joelle Pineau · Mike Rabbat Multi-simulator training has contributed to the recent success of Deep Reinforcement Learning (Deep RL) by stabilizing learning and allowing for higher training throughputs. In this work, we propose Gossip-based Actor-Learner Architectures (GALA) where several actor-learners (such as A2C agents) are organized in a peer-to-peer communication topology, and exchange information through asynchronous gossip in order to take advantage of a large number of distributed simulators. We prove that GALA agents remain within an epsilon-ball of one-another during training when using loosely coupled asynchronous communication. By reducing the amount of synchronization between agents, GALA is more computationally efficient and scalable compared to A2C, its fully-synchronous counterpart. GALA also outperforms A2C, being more robust and sample efficient. We show that we can run several loosely coupled GALA agents in parallel on a single GPU and achieve significantly higher hardware utilization and frame-rates than vanilla A2C at comparable power draws.

# Exploration and Application

## Explicit Planning for Efficient Exploration in Reinforcement Learning

Liangpeng Zhang · Ke Tang · Xin Yao Efficient exploration is crucial to achieving good performance in reinforcement learning. Existing systematic exploration strategies (R-MAX, MBIE, UCRL, etc.), despite being promising theoretically, are essentially greedy strategies that follow some predefined heuristics. When the heuristics do not match the dynamics of Markov decision processes (MDPs) well, an excessive amount of time can be wasted in travelling through already-explored states, lowering the overall efficiency. We argue that explicit planning for exploration can help alleviate such a problem, and propose a Value Iteration for Exploration Cost (VIEC) algorithm which computes the optimal exploration scheme by solving an augmented MDP. We then present a detailed analysis of the exploration behaviour of some popular strategies, showing how these strategies can fail and spend O(n^2 md) or O(n^2 m + nmd) steps to collect sufficient data in some tower-shaped MDPs, while the optimal exploration scheme, which can be obtained by VIEC, only needs O(nmd), where n, m are the numbers of states and actions and d is the data demand. The analysis not only points out the weakness of existing heuristic-based strategies, but also suggests a remarkable potential in explicit planning for exploration.

# Meta Learning

## *Unsupervised Curricula for Visual Meta-Reinforcement Learning

Allan Jabri · Kyle Hsu · Abhishek Gupta · Ben Eysenbach · Sergey Levine · Chelsea Finn In principle, meta-reinforcement learning algorithms leverage experience across many tasks to learn fast and effective reinforcement learning (RL) strategies. However, current meta-RL approaches rely on manually-defined distributions of training tasks, and hand-crafting these task distributions can be challenging and time-consuming. Can useful’’ pre-training tasks be discovered in an unsupervised manner? We develop an unsupervised algorithm for inducing an adaptive meta-training task distribution, i.e. an automatic curriculum, by modeling unsupervised interaction in a visual environment. The task distribution is scaffolded by a parametric density model of the meta-learner’s trajectory distribution. We formulate unsupervised meta-RL as information maximization between a latent task variable and the meta-learner’s data distribution, and describe a practical instantiation which alternates between integration of recent experience into the task distribution and meta-learning of the updated tasks. Repeating this procedure leads to iterative reorganization such that the curriculum adapts as the meta-learner’s data distribution shifts. Moreover, we show how discriminative clustering frameworks for visual representations can support trajectory-level task acquisition and exploration in domains with pixel observations, avoiding the pitfalls of alternatives. In experiments on vision-based navigation and manipulation domains, we show that the algorithm allows for unsupervised meta-learning that both transfers to downstream tasks specified by hand-crafted reward functions and serves as pre-training for more efficient meta-learning of test task distributions.

## A Meta-MDP Approach to Exploration for Lifelong Reinforcement Learning

Francisco Garcia · Philip Thomas In this paper we consider the problem of how a reinforcement learning agent that is tasked with solving a sequence of reinforcement learning problems (a sequence of Markov decision processes) can use knowledge acquired early in its lifetime to improve its ability to solve new problems. We argue that previous experience with similar problems can provide an agent with information about how it should explore when facing a new but related problem. We show that the search for an optimal exploration strategy can be formulated as a reinforcement learning problem itself and demonstrate that such strategy can leverage patterns found in the structure of related problems. We conclude with experiments that show the benefits of optimizing an exploration strategy using our proposed framework.

## SMILe: Scalable Meta Inverse Reinforcement Learning through Context-Conditional Policies

Seyed Kamyar Seyed Ghasemipour · Shixiang (Shane) Gu · Richard Zemel Imitation Learning (IL) has been successfully applied to complex sequential decision-making problems where standard Reinforcement Learning (RL) algorithms fail. A number of recent methods extend IL to few-shot learning scenarios, where a meta-trained policy learns to quickly master new tasks using limited demonstrations. However, although Inverse Reinforcement Learning (IRL) often outperforms Behavioral Cloning (BC) in terms of imitation quality, most of these approaches build on BC due to its simple optimization objective. In this work, we propose SMILe, a scalable framework for Meta Inverse Reinforcement Learning (Meta-IRL) based on maximum entropy IRL, which can learn high-quality policies from few demonstrations. We examine the efficacy of our method on a variety of high-dimensional simulated continuous control tasks and observe that SMILe significantly outperforms Meta-BC. Furthermore, we observe that SMILe performs comparably or outperforms Meta-DAgger, while being applicable in the state-only setting and not requiring online experts. To our knowledge, our approach is the first efficient method for Meta-IRL that scales to the function approximator setting. For datasets and reproducing results please refer to https://github.com/KamyarGh/rlswiss/blob/master/reproducing/smilepaper.md .

## Meta-Inverse Reinforcement Learning with Probabilistic Context Variables

Lantao Yu · Tianhe Yu · Chelsea Finn · Stefano Ermon Reinforcement learning demands a reward function, which is often difficult to provide or design in real world applications. While inverse reinforcement learning (IRL) holds promise for automatically learning reward functions from demonstrations, several major challenges remain. First, existing IRL methods learn reward functions from scratch, requiring large numbers of demonstrations to correctly infer the reward for each task the agent may need to perform. Second, and more subtly, existing methods typically assume demonstrations for one, isolated behavior or task, while in practice, it is significantly more natural and scalable to provide datasets of heterogeneous behaviors. To this end, we propose a deep latent variable model that is capable of learning rewards from unstructured, multi-task demonstration data, and critically, use this experience to infer robust rewards for new, structurally-similar tasks from a single demonstration. Our experiments on multiple continuous control tasks demonstrate the effectiveness of our approach compared to state-of-the-art imitation and inverse reinforcement learning methods.

# Hierarchical Reinforcement Learning

## Hierarchical Reinforcement Learning with Advantage-Based Auxiliary Rewards

Siyuan Li · Rui Wang · Minxue Tang · Chongjie Zhang 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, simultaneous learning of the high-level policy and low-level skills without using task-specific knowledge. In addition, we also theoretically prove that optimizing low-level skills with this auxiliary reward will increase the task return for the joint policy. Experimental results show that our algorithm dramatically outperforms other state-of-the-art HRL methods in Mujoco domains. We also find both low-level and high-level policies trained by our algorithm transferable.

## Language as an Abstraction for Hierarchical Deep Reinforcement Learning

YiDing Jiang · Shixiang (Shane) Gu · Kevin Murphy · Chelsea Finn Solving complex, temporally-extended tasks is a long-standing problem in reinforcement learning (RL). We hypothesize that one critical element of solving such problems is the notion of compositionality. With the ability to learn sub-skills that can be composed to solve longer tasks, i.e. hierarchical RL, we can acquire temporally-extended behaviors. However, acquiring effective yet general abstractions for hierarchical RL is remarkably challenging. In this paper, we propose to use language as the abstraction, as it provides unique compositional structure, enabling fast learning and combinatorial generalization, while retaining tremendous flexibility, making it suitable for a variety of problems. Our approach learns an instruction-following low-level policy and a high-level policy that can reuse abstractions across tasks, in essence, permitting agents to reason using structured language. To study compositional task learning, we introduce an open-source object interaction environment built using the MuJoCo physics engine and the CLEVR engine. We find that, using our approach, agents can learn to solve to diverse, temporally-extended tasks such as object sorting and multi-object rearrangement, including from raw pixel observations. Our analysis find that the compositional nature of language is critical for learning and systematically generalizing sub-skills in comparison to non-compositional abstractions that use the same supervision.

# Inversed Reinforcement Learning

## Learner-aware Teaching: Inverse Reinforcement Learning with Preferences and Constraints

Sebastian Tschiatschek · Ahana Ghosh · Luis Haug · Rati Devidze · Adish Singla Inverse reinforcement learning (IRL) enables an agent to learn complex behavior by observing demonstrations from a (near-)optimal policy. The typical assumption is that the learner’s goal is to match the teacher’s demonstrated behavior. In this paper, we consider the setting where the learner has its own preferences that it additionally takes into consideration. These preferences can for example capture behavioral biases, mismatched worldviews, or physical constraints. We study two teaching approaches: learner-agnostic teaching, where the teacher provides demonstrations from an optimal policy ignoring the learner’s preferences, and learner-aware teaching, where the teacher accounts for the learner’s preferences. We design learner-aware teaching algorithms and show that significant performance improvements can be achieved over learner-agnostic teaching.

## On the Correctness and Sample Complexity of Inverse Reinforcement Learning

Abi Komanduru · Jean Honorio Inverse reinforcement learning (IRL) is the problem of finding a reward function that generates a given optimal policy for a given Markov Decision Process. This paper looks at an algorithmic-independent geometric analysis of the IRL problem with finite states and actions. A L1-regularized Support Vector Machine formulation of the IRL problem motivated by the geometric analysis is then proposed with the basic objective of the inverse reinforcement problem in mind: to find a reward function that generates a specified optimal policy. The paper further analyzes the proposed formulation of inverse reinforcement learning with n states and k actions, and shows a sample complexity of O(d2log(nk)) for transition probability matrices with at most d non-zeros per row, for recovering a reward function that generates a policy that satisfies Bellman’s optimality condition with respect to the true transition probabilities.

# Multiple Objectives/Agents

## Value Propagation for Decentralized Networked Deep Multi-agent Reinforcement Learning

Chao Qu · Shie Mannor · Huan Xu · Yuan Qi · Le Song · Junwu Xiong We consider the networked multi-agent reinforcement learning (MARL) problem in a fully decentralized setting, where agents learn to coordinate to achieve joint success. This problem is widely encountered in many areas including traffic control, distributed control, and smart grids. We assume each agent is located at a node of a communication network and can exchange information only with its neighbors. Using softmax temporal consistency, we derive a primal-dual decentralized optimization method and obtain a principled and data-efficient iterative algorithm named {\em value propagation}. We prove a non-asymptotic convergence rate of O(1/T) with nonlinear function approximation. To the best of our knowledge, it is the first MARL algorithm with a convergence guarantee in the control, off-policy, non-linear function approximation, fully decentralized setting.

## Efficient Communication in Multi-Agent Reinforcement Learning via Variance Based Control

Sai Qian Zhang · Qi Zhang · Jieyu Lin Multi-agent reinforcement learning (MARL) has recently received considerable attention due to its applicability to a wide range of real-world applications. However, achieving efficient communication among agents has always been an overarching problem in MARL. In this work, we propose Variance Based Control (VBC), a simple yet efficient technique to improve communication efficiency in MARL. By limiting the variance of the exchanged messages between agents during the training phase, the noisy component in the messages can be eliminated effectively, while the useful part can be preserved and utilized by the agents for better performance. Our evaluation using multiple MARL benchmarks indicates that our method achieves 2−10× lower in communication overhead than state-of-the-art MARL algorithms, while allowing agents to achieve better overall performance.

## LIIR: Learning Individual Intrinsic Reward in Multi-Agent Reinforcement Learning

Yali Du · Lei Han · Meng Fang · Ji Liu · Tianhong Dai · Dacheng Tao A great challenge in cooperative decentralized multi-agent reinforcement learning (MARL) is generating diversified behaviors for each individual agent when receiving only a team reward. Prior studies have paid much effort on reward shaping or designing a centralized critic that can discriminatively credit the agents. In this paper, we propose to merge the two directions and learn each agent an intrinsic reward function which diversely stimulates the agents at each time step. Specifically, the intrinsic reward for a specific agent will be involved in computing a distinct proxy critic for the agent to direct the updating of its individual policy. Meanwhile, the parameterized intrinsic reward function will be updated towards maximizing the expected accumulated team reward from the environment so that the objective is consistent with the original MARL problem. The proposed method is referred to as learning individual intrinsic reward (LIIR) in MARL. We compare LIIR with a number of state-of-the-art MARL methods on battle games in StarCraft II. The results demonstrate the effectiveness of LIIR, and we show LIIR can assign each individual agent an insightful intrinsic reward per time step.

## A Structured Prediction Approach for Generalization in Cooperative Multi-Agent Reinforcement Learning

Nicolas Carion · Nicolas Usunier · Gabriel Synnaeve · Alessandro Lazaric Effective coordination is crucial to solve multi-agent collaborative (MAC) problems. While centralized reinforcement learning methods can optimally solve small MAC instances, they do not scale to large problems and they fail to generalize to scenarios different from those seen during training. In this paper, we consider MAC problems with some intrinsic notion of locality (e.g., geographic proximity) such that interactions between agents and tasks are locally limited. By leveraging this property, we introduce a novel structured prediction approach to assign agents to tasks. At each step, the assignment is obtained by solving a centralized optimization problem (the inference procedure) whose objective function is parameterized by a learned scoring model. We propose different combinations of inference procedures and scoring models able to represent coordination patterns of increasing complexity. The resulting assignment policy can be efficiently learned on small problem instances and readily reused in problems with more agents and tasks (i.e., zero-shot generalization). We report experimental results on a toy search and rescue problem and on several target selection scenarios in StarCraft: Brood War, in which our model significantly outperforms strong rule-based baselines on instances with 5 times more agents and tasks than those seen during training.

## Multi-Agent Common Knowledge Reinforcement Learning

Christian Schroeder de Witt · Jakob Foerster · Gregory Farquhar · Philip Torr · Wendelin Boehmer · Shimon Whiteson Cooperative multi-agent reinforcement learning often requires decentralised policies, which severely limit the agents’ ability to coordinate their behaviour. In this paper, we show that common knowledge between agents allows for complex decentralised coordination. Common knowledge arises naturally in a large number of decentralised cooperative multi-agent tasks, for example, when agents can reconstruct parts of each others’ observations. Since agents can independently agree on their common knowledge, they can execute complex coordinated policies that condition on this knowledge in a fully decentralised fashion. We propose multi-agent common knowledge reinforcement learning (MACKRL), a novel stochastic actor-critic algorithm that learns a hierarchical policy tree. Higher levels in the hierarchy coordinate groups of agents by conditioning on their common knowledge, or delegate to lower levels with smaller subgroups but potentially richer common knowledge. The entire policy tree can be executed in a fully decentralised fashion. As the lowest policy tree level consists of independent policies for each agent, MACKRL reduces to independently learnt decentralised policies as a special case. We demonstrate that our method can exploit common knowledge for superior performance on complex decentralised coordination tasks, including a stochastic matrix game and challenging problems in StarCraft II unit micromanagement.

## Biases for Emergent Communication in Multi-agent Reinforcement Learning

Tom Eccles · Yoram Bachrach · Guy Lever · Angeliki Lazaridou · Thore Graepel We study the problem of emergent communication, in which language arises because speakers and listeners must communicate information in order to solve tasks. In temporally extended reinforcement learning domains, it has proved hard to learn such communication without centralized training of agents, due in part to a difficult joint exploration problem. We introduce inductive biases for positive signalling and positive listening, which ease this problem. In a simple one-step environment, we demonstrate how these biases ease the learning problem. We also apply our methods to a more extended environment, showing that agents with these inductive biases achieve better performance, and analyse the resulting communications protocols.

# Reward Function

## Distributional Reward Decomposition for Reinforcement Learning

Zichuan Lin · Li Zhao · Derek Yang · Tao Qin · Tie-Yan Liu · Guangwen Yang Many reinforcement learning (RL) tasks have specific properties that can be leveraged to modify existing RL algorithms to adapt to those tasks and further improve performance, and a general class of such properties is the multiple reward channel. In those environments the full reward can be decomposed into sub-rewards obtained from different channels. Existing work on reward decomposition either requires prior knowledge of the environment to decompose the full reward, or decomposes reward without prior knowledge but with degraded performance. In this paper, we propose Distributional Reward Decomposition for Reinforcement Learning (DRDRL), a novel reward decomposition algorithm which captures the multiple reward channel structure under distributional setting. Empirically, our method captures the multi-channel structure and discovers meaningful reward decomposition, without any requirements on prior knowledge. Consequently, our agent achieves better performance than existing methods on environments with multiple reward channels.

## Learning Reward Machines for Partially Observable Reinforcement Learning

Rodrigo Toro Icarte · Ethan Waldie · Toryn Klassen · Rick Valenzano · Margarita Castro · Sheila McIlraith Reward Machines (RMs), originally proposed for specifying problems in Reinforcement Learning (RL), provide a structured, automata-based representation of a reward function that allows an agent to decompose problems into subproblems that can be efficiently learned using off-policy learning. Here we show that RMs can be learned from experience, instead of being specified by the user, and that the resulting problem decomposition can be used to effectively solve partially observable RL problems. We pose the task of learning RMs as a discrete optimization problem where the objective is to find an RM that decomposes the problem into a set of subproblems such that the combination of their optimal memoryless policies is an optimal policy for the original problem. We show the effectiveness of this approach on three partially observable domains, where it significantly outperforms A3C, PPO, and ACER, and discuss its advantages, limitations, and broader potential.

# Applications

## Staying up to Date with Online Content Changes Using Reinforcement Learning for Scheduling

Andrey Kolobov · Yuval Peres · Cheng Lu · Eric Horvitz From traditional Web search engines to virtual assistants and Web accelerators, services that rely on online information need to continually keep track of remote content changes by explicitly requesting content updates from remote sources (e.g., web pages). We propose a novel optimization objective for this setting that has several practically desirable properties, and efficient algorithms for it with optimality guarantees even in the face of mixed content change observability and initially unknown change model parameters. Experiments on 18.5M URLs crawled daily for 14 weeks show significant advantages of this approach over prior art.

## Loaded DiCE: Trading off Bias and Variance in Any-Order Score Function Gradient Estimators for Reinforcement Learning

Gregory Farquhar · Shimon Whiteson · Jakob Foerster Gradient-based methods for optimisation of objectives in stochastic settings with unknown or intractable dynamics require estimators of derivatives. We derive an objective that, under automatic differentiation, produces low-variance unbiased estimators of derivatives at any order. Our objective is compatible with arbitrary advantage estimators, which allows the control of the bias and variance of any-order derivatives when using function approximation. Furthermore, we propose a method to trade off bias and variance of higher order derivatives by discounting the impact of more distant causal dependencies. We demonstrate the correctness and utility of our estimator in analytically tractable MDPs and in meta-reinforcement-learning for continuous control.

## A Composable Specification Language for Reinforcement Learning Tasks

Kishor Jothimurugan · Rajeev Alur · Osbert Bastani Reinforcement learning is a promising approach for learning control policies for robot tasks. However, specifying complex tasks (e.g., with multiple objectives and safety constraints) can be challenging, since the user must design a reward function that encodes the entire task. Furthermore, the user often needs to manually shape the reward to ensure convergence of the learning algorithm. We propose a language for specifying complex control tasks, along with an algorithm that compiles specifications in our language into a reward function and automatically performs reward shaping. We implement our approach in a tool called SPECTRL, and show that it outperforms several state-of-the-art baselines.

## Near-Optimal Reinforcement Learning in Dynamic Treatment Regimes

Junzhe Zhang · Elias Bareinboim A dynamic treatment regime (DTR) consists of a sequence of decision rules, one per stage of intervention, that dictates how to determine the treatment assignment to patients based on evolving treatments and covariates’ history. These regimes are particularly effective for managing chronic disorders and is arguably one of the key aspects towards more personalized decision-making. In this paper, we investigate the online reinforcement learning (RL) problem for selecting optimal DTRs provided that observational data is available. We develop the first adaptive algorithm that achieves near-optimal regret in DTRs in online settings, without any access to historical data. We further derive informative bounds on the system dynamics of the underlying DTR from confounded, observational data. Finally, we combine these results and develop a novel RL algorithm that efficiently learns the optimal DTR while leveraging the abundant, yet imperfect confounded observations.

# Other

## Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck

Maximilian Igl · Kamil Ciosek · Yingzhen Li · Sebastian Tschiatschek · Cheng Zhang · Sam Devlin · Katja Hofmann The ability for policies to generalize to new environments is key to the broad application of RL agents. A promising approach to prevent an agent’s policy from overfitting to a limited set of training environments is to apply regularization techniques originally developed for supervised learning. However, there are stark differences between supervised learning and RL. We discuss those differences and propose modifications to existing regularization techniques in order to better adapt them to RL. In particular, we focus on regularization techniques relying on the injection of noise into the learned function, a family that includes some of the most widely used approaches such as Dropout and Batch Normalization. To adapt them to RL, we propose Selective Noise Injection (SNI), which maintains the regularizing effect the injected noise has, while mitigating the adverse effects it has on the gradient quality. Furthermore, we demonstrate that the Information Bottleneck (IB) is a particularly well suited regularization technique for RL as it is effective in the low-data regime encountered early on in training RL agents. Combining the IB with SNI, we significantly outperform current state of the art results, including on the recently proposed generalization benchmark Coinrun.

## Using a Logarithmic Mapping to Enable Lower Discount Factors in Reinforcement Learning

Harm Van Seijen · Mehdi Fatemi · Arash Tavakoli In an effort to better understand the different ways in which the discount factor affects the optimization process in reinforcement learning, we designed a set of experiments to study each effect in isolation. Our analysis reveals that the common perception that poor performance of low discount factors is caused by (too) small action-gaps requires revision. We propose an alternative hypothesis that identifies the size-difference of the action-gap across the state-space as the primary cause. We then introduce a new method that enables more homogeneous action-gaps by mapping value estimates to a logarithmic space. We prove convergence for this method under standard assumptions and demonstrate empirically that it indeed enables lower discount factors for approximate reinforcement-learning methods. This in turn allows tackling a class of reinforcement-learning problems that are challenging to solve with traditional methods.

## Search on the Replay Buffer: Bridging Planning and Reinforcement Learning

Ben Eysenbach · Russ Salakhutdinov · Sergey Levine The history of learning for control has been an exciting back and forth between two broad classes of algorithms: planning and reinforcement learning. Planning algorithms effectively reason over long horizons, but assume access to a local policy and distance metric over collision-free paths. Reinforcement learning excels at learning policies and relative values of states, but fails to plan over long horizons. Despite the successes of each method on various tasks, long horizon, sparse reward tasks with high-dimensional observations remain exceedingly challenging for both planning and reinforcement learning algorithms. Frustratingly, these sorts of tasks are potentially the most useful, as they are simple to design (a human only need to provide an example goal state) and avoid injecting bias through reward shaping. We introduce a general-purpose control algorithm that combines the strengths of planning and reinforcement learning to effectively solve these tasks. Our main idea is to decompose the task of reaching a distant goal state into a sequence of easier tasks, each of which corresponds to reaching a particular subgoal. We use goal-conditioned RL to learn a policy to reach each waypoint and to learn a distance metric for search. Using graph search over our replay buffer, we can automatically generate this sequence of subgoals, even in image-based environments. Our algorithm, search on the replay buffer (SoRB), enables agents to solve sparse reward tasks over hundreds of steps, and generalizes substantially better than standard RL algorithms.

## [A Family of Robust Stochastic Operators for Reinforcement Learning

](https://nips.cc/Conferences/2019/Schedule?showEvent=14453)

Yingdong Lu · Mark Squillante · Chai Wah Wu We consider a new family of stochastic operators for reinforcement learning with the goal of alleviating negative effects and becoming more robust to approximation or estimation errors. Various theoretical results are established, which include showing that our family of operators preserve optimality and increase the action gap in a stochastic sense. Our empirical results illustrate the strong benefits of our robust stochastic operators, significantly outperforming the classical Bellman operator and recently proposed operators.