On the Interplay Between Sparsity and Training in Deep Reinforcement Learning

no code implementations • 28 Jan 2025 • Fatima Davelouis, John D. Martin, Michael Bowling

We study the benefits of different sparse architectures for deep reinforcement learning.

Deep Reinforcement Learning reinforcement-learning

Meta-Gradient Search Control: A Method for Improving the Efficiency of Dyna-style Planning

no code implementations • 27 Jun 2024 • Bradley Burega, John D. Martin, Luke Kapeluck, Michael Bowling

We study how a Reinforcement Learning (RL) system can remain sample-efficient when learning from an imperfect model of the environment.

Reinforcement Learning (RL)

MOTO: Offline Pre-training to Online Fine-tuning for Model-based Robot Learning

no code implementations • 6 Jan 2024 • Rafael Rafailov, Kyle Hatch, Victor Kolev, John D. Martin, Mariano Phielipp, Chelsea Finn

We study the problem of offline pre-training and online fine-tuning for reinforcement learning from high-dimensional observations in the context of realistic robot tasks.

Offline RL Robot Manipulation

Settling the Reward Hypothesis

no code implementations • 20 Dec 2022 • Michael Bowling, John D. Martin, David Abel, Will Dabney

The reward hypothesis posits that, "all of what we mean by goals and purposes can be well thought of as maximization of the expected value of the cumulative sum of a received scalar signal (reward)."

Should Models Be Accurate?

no code implementations • 22 May 2022 • Esra'a Saleh, John D. Martin, Anna Koop, Arash Pourzarabi, Michael Bowling

We focus our investigations on Dyna-style planning in a prediction setting.

Meta-Learning Model-based Reinforcement Learning

Adapting the Function Approximation Architecture in Online Reinforcement Learning

no code implementations • 17 Jun 2021 • John D. Martin, Joseph Modayil

However, prevailing optimization techniques are not designed for strictly-incremental online updates.

reinforcement-learning Reinforcement Learning +1

Variational Filtering with Copula Models for SLAM

no code implementations • 2 Aug 2020 • John D. Martin, Kevin Doherty, Caralyn Cyr, Brendan Englot, John Leonard

The ability to infer map variables and estimate pose is crucial to the operation of autonomous mobile robots.

Simultaneous Localization and Mapping

Autonomous Exploration Under Uncertainty via Deep Reinforcement Learning on Graphs

1 code implementation • 24 Jul 2020 • Fanfei Chen, John D. Martin, Yewei Huang, Jinkun Wang, Brendan Englot

We consider an autonomous exploration problem in which a range-sensing mobile robot is tasked with accurately mapping the landmarks in an a priori unknown environment efficiently in real-time; it must choose sensing actions that both curb localization uncertainty and achieve information gain.

Decision Making Deep Reinforcement Learning +2

On Catastrophic Interference in Atari 2600 Games

1 code implementation • 28 Feb 2020 • William Fedus, Dibya Ghosh, John D. Martin, Marc G. Bellemare, Yoshua Bengio, Hugo Larochelle

Our study provides a clear empirical link between catastrophic interference and sample efficiency in reinforcement learning.

Atari Games Deep Reinforcement Learning +2

Stochastically Dominant Distributional Reinforcement Learning

no code implementations • ICML 2020 • John D. Martin, Michal Lyskawinski, Xiaohu Li, Brendan Englot

We describe a new approach for managing aleatoric uncertainty in the Reinforcement Learning (RL) paradigm.

Distributional Reinforcement Learning reinforcement-learning +2