Marco Pleines

Marco Pleines, Matthias Pallasch, Frank Zimmer et al.

At first sight it may seem straightforward to use recurrent layers in Deep Reinforcement Learning algorithms to enable agents to make use of memory in the setting of partially observable environments. Starting from widely used Proximal Policy Optimization (PPO), we highlight vital details that one must get right when adding recurrence to achieve a correct and efficient implementation, namely: properly shaping the neural net's forward pass, arranging the training data, correspondingly selecting hidden states for sequence beginnings and masking paddings for loss computation. We further explore the limitations of recurrent PPO by benchmarking the contributed novel environments Mortar Mayhem and Searing Spotlights that challenge the agent's memory beyond solely capacity and distraction tasks. Remarkably, we can demonstrate a transition to strong generalization in Mortar Mayhem when scaling the number of training seeds, while the agent does not succeed on Searing Spotlights, which seems to be a tough challenge for memory-based agents.

Marco Pleines, Matthias Pallasch, Frank Zimmer et al.

Memory Gym presents a suite of 2D partially observable environments, namely Mortar Mayhem, Mystery Path, and Searing Spotlights, designed to benchmark memory capabilities in decision-making agents. These environments, originally with finite tasks, are expanded into innovative, endless formats, mirroring the escalating challenges of cumulative memory games such as "I packed my bag". This progression in task design shifts the focus from merely assessing sample efficiency to also probing the levels of memory effectiveness in dynamic, prolonged scenarios. To address the gap in available memory-based Deep Reinforcement Learning baselines, we introduce an implementation within the open-source CleanRL library that integrates Transformer-XL (TrXL) with Proximal Policy Optimization. This approach utilizes TrXL as a form of episodic memory, employing a sliding window technique. Our comparative study between the Gated Recurrent Unit (GRU) and TrXL reveals varied performances across our finite and endless tasks. TrXL, on the finite environments, demonstrates superior effectiveness over GRU, but only when utilizing an auxiliary loss to reconstruct observations. Notably, GRU makes a remarkable resurgence in all endless tasks, consistently outperforming TrXL by significant margins. Website and Source Code: https://marcometer.github.io/jmlr_2024.github.io/

Marco Pleines, Konstantin Ramthun, Yannik Wegener et al.

Autonomously trained agents that are supposed to play video games reasonably well rely either on fast simulation speeds or heavy parallelization across thousands of machines running concurrently. This work explores a third way that is established in robotics, namely sim-to-real transfer, or if the game is considered a simulation itself, sim-to-sim transfer. In the case of Rocket League, we demonstrate that single behaviors of goalies and strikers can be successfully learned using Deep Reinforcement Learning in the simulation environment and transferred back to the original game. Although the implemented training simulation is to some extent inaccurate, the goalkeeping agent saves nearly 100% of its faced shots once transferred, while the striking agent scores in about 75% of cases. Therefore, the trained agent is robust enough and able to generalize to the target domain of Rocket League.

Marco Pleines, Daniel Addis, David Rubinstein et al.

Pokémon Red, a classic Game Boy JRPG, presents significant challenges as a testbed for agents, including multi-tasking, long horizons of tens of thousands of steps, hard exploration, and a vast array of potential policies. We introduce a simplistic environment and a Deep Reinforcement Learning (DRL) training methodology, demonstrating a baseline agent that completes an initial segment of the game up to completing Cerulean City. Our experiments include various ablations that reveal vulnerabilities in reward shaping, where agents exploit specific reward signals. We also discuss limitations and argue that games like Pokémon hold strong potential for future research on Large Language Model agents, hierarchical training algorithms, and advanced exploration methods. Source Code: https://github.com/MarcoMeter/neroRL/tree/poke_red

Jonas Schumacher, Marco Pleines

In this work, the trick-taking game Wizard with a separate bidding and playing phase is modeled by two interleaved partially observable Markov decision processes (POMDP). Deep Q-Networks (DQN) are used to empower self-improving agents, which are capable of tackling the challenges of a highly non-stationary environment. To compare algorithms between each other, the accuracy between bid and trick count is monitored, which strongly correlates with the actual rewards and provides a well-defined upper and lower performance bound. The trained DQN agents achieve accuracies between 66% and 87% in self-play, leaving behind both a random baseline and a rule-based heuristic. The conducted analysis also reveals a strong information asymmetry concerning player positions during bidding. To overcome the missing Markov property of imperfect-information games, a long short-term memory (LSTM) network is implemented to integrate historic information into the decision-making process. Additionally, a forward-directed tree search is conducted by sampling a state of the environment and thereby turning the game into a perfect information setting. To our surprise, both approaches do not surpass the performance of the basic DQN agent.

Marco Pleines, Jenia Jitsev, Mike Preuss et al.

The Obstacle Tower Challenge is the task to master a procedurally generated chain of levels that subsequently get harder to complete. Whereas the most top performing entries of last year's competition used human demonstrations or reward shaping to learn how to cope with the challenge, we present an approach that performed competitively (placed 7th) but starts completely from scratch by means of Deep Reinforcement Learning with a relatively simple feed-forward deep network structure. We especially look at the generalization performance of the taken approach concerning different seeds and various visual themes that have become available after the competition, and investigate where the agent fails and why. Note that our approach does not possess a short-term memory like employing recurrent hidden states. With this work, we hope to contribute to a better understanding of what is possible with a relatively simple, flexible solution that can be applied to learning in environments featuring complex 3D visual input where the abstract task structure itself is still fairly simple.