Enhanced Generalization through Prioritization and Diversity in Self-Imitation Reinforcement Learning over Procedural Environments with Sparse Rewards

Exploration poses a fundamental challenge in Reinforcement Learning (RL) with sparse rewards, limiting an agent's ability to learn optimal decision-making due to a lack of informative feedback signals. Self-Imitation Learning (self-IL) has emerged as a promising approach for exploration, leveraging a replay buffer to store and reproduce successful behaviors. However, traditional self-IL methods, which rely on high-return transitions and assume singleton environments, face challenges in generalization, especially in procedurally-generated (PCG) environments. Therefore, new self-IL methods have been proposed to rank which experiences to persist, but they replay transitions uniformly regardless of their significance, and do not address the diversity of the stored demonstrations. In this work, we propose tailored self-IL sampling strategies by prioritizing transitions in different ways and extending prioritization techniques to PCG environments. We also address diversity loss through modifications to counteract the impact of generalization requirements and bias introduced by prioritization techniques. Our experimental analysis, conducted over three PCG sparse reward environments, including MiniGrid and ProcGen, highlights the benefits of our proposed modifications, achieving a new state-of-the-art performance in the MiniGrid-MultiRoom-N12-S10 environment.

Using Offline Data to Speed-up Reinforcement Learning in Procedurally Generated Environments

One of the key challenges of Reinforcement Learning (RL) is the ability of agents to generalise their learned policy to unseen settings. Moreover, training RL agents requires large numbers of interactions with the environment. Motivated by the recent success of Offline RL and Imitation Learning (IL), we conduct a study to investigate whether agents can leverage offline data in the form of trajectories to improve the sample-efficiency in procedurally generated environments. We consider two settings of using IL from offline data for RL: (1) pre-training a policy before online RL training and (2) concurrently training a policy with online RL and IL from offline data. We analyse the impact of the quality (optimality of trajectories) and diversity (number of trajectories and covered level) of available offline trajectories on the effectiveness of both approaches. Across four well-known sparse reward tasks in the MiniGrid environment, we find that using IL for pre-training and concurrently during online RL training both consistently improve the sample-efficiency while converging to optimal policies. Furthermore, we show that pre-training a policy from as few as two trajectories can make the difference between learning an optimal policy at the end of online training and not learning at all. Our findings motivate the widespread adoption of IL for pre-training and concurrent IL in procedurally generated environments whenever offline trajectories are available or can be generated.

Towards Improving Exploration in Self-Imitation Learning using Intrinsic Motivation

Reinforcement Learning has emerged as a strong alternative to solve optimization tasks efficiently. The use of these algorithms highly depends on the feedback signals provided by the environment in charge of informing about how good (or bad) the decisions made by the learned agent are. Unfortunately, in a broad range of problems the design of a good reward function is not trivial, so in such cases sparse reward signals are instead adopted. The lack of a dense reward function poses new challenges, mostly related to exploration. Imitation Learning has addressed those problems by leveraging demonstrations from experts. In the absence of an expert (and its subsequent demonstrations), an option is to prioritize well-suited exploration experiences collected by the agent in order to bootstrap its learning process with good exploration behaviors. However, this solution highly depends on the ability of the agent to discover such trajectories in the early stages of its learning process. To tackle this issue, we propose to combine imitation learning with intrinsic motivation, two of the most widely adopted techniques to address problems with sparse reward. In this work intrinsic motivation is used to encourage the agent to explore the environment based on its curiosity, whereas imitation learning allows repeating the most promising experiences to accelerate the learning process. This combination is shown to yield an improved performance and better generalization in procedurally-generated environments, outperforming previously reported self-imitation learning methods and achieving equal or better sample efficiency with respect to intrinsic motivation in isolation.

An Evaluation Study of Intrinsic Motivation Techniques applied to Reinforcement Learning over Hard Exploration Environments

In the last few years, the research activity around reinforcement learning tasks formulated over environments with sparse rewards has been especially notable. Among the numerous approaches proposed to deal with these hard exploration problems, intrinsic motivation mechanisms are arguably among the most studied alternatives to date. Advances reported in this area over time have tackled the exploration issue by proposing new algorithmic ideas to generate alternative mechanisms to measure the novelty. However, most efforts in this direction have overlooked the influence of different design choices and parameter settings that have also been introduced to improve the effect of the generated intrinsic bonus, forgetting the application of those choices to other intrinsic motivation techniques that may also benefit of them. Furthermore, some of those intrinsic methods are applied with different base reinforcement algorithms (e.g. PPO, IMPALA) and neural network architectures, being hard to fairly compare the provided results and the actual progress provided by each solution. The goal of this work is to stress on this crucial matter in reinforcement learning over hard exploration environments, exposing the variability and susceptibility of avant-garde intrinsic motivation techniques to diverse design factors. Ultimately, our experiments herein reported underscore the importance of a careful selection of these design aspects coupled with the exploration requirements of the environment and the task in question under the same setup, so that fair comparisons can be guaranteed.

Collaborative Training of Heterogeneous Reinforcement Learning Agents in Environments with Sparse Rewards: What and When to Share?

In the early stages of human life, babies develop their skills by exploring different scenarios motivated by their inherent satisfaction rather than by extrinsic rewards from the environment. This behavior, referred to as intrinsic motivation, has emerged as one solution to address the exploration challenge derived from reinforcement learning environments with sparse rewards. Diverse exploration approaches have been proposed to accelerate the learning process over single- and multi-agent problems with homogeneous agents. However, scarce studies have elaborated on collaborative learning frameworks between heterogeneous agents deployed into the same environment, but interacting with different instances of the latter without any prior knowledge. Beyond the heterogeneity, each agent's characteristics grant access only to a subset of the full state space, which may hide different exploration strategies and optimal solutions. In this work we combine ideas from intrinsic motivation and transfer learning. Specifically, we focus on sharing parameters in actor-critic model architectures and on combining information obtained through intrinsic motivation with the aim of having a more efficient exploration and faster learning. We test our strategies through experiments performed over a modified ViZDooM's My Way Home scenario, which is more challenging than its original version and allows evaluating the heterogeneity between agents. Our results reveal different ways in which a collaborative framework with little additional computational cost can outperform an independent learning process without knowledge sharing. Additionally, we depict the need for modulating correctly the importance between the extrinsic and intrinsic rewards to avoid undesired agent behaviors.