Adrien Ali Taiga

Yevgen Chebotar

Ted Xiao

Alex Irpan

Sergey Levine

Pablo Samuel Castro

Aleksandra Faust

Aviral Kumar

Rishabh Agarwal

Value functions are a central component of deep reinforcement learning (RL). These functions, parameterized by neural networks, are trained … (voir plus)using a mean squared error regression objective to match bootstrapped target values. However, scaling value-based RL methods that use regression to large networks, such as high-capacity Transformers, has proven challenging. This difficulty is in stark contrast to supervised learning: by leveraging a cross-entropy classification loss, supervised methods have scaled reliably to massive networks. Observing this discrepancy, in this paper, we investigate whether the scalability of deep RL can also be improved simply by using classification in place of regression for training value functions. We demonstrate that value functions trained with categorical cross-entropy significantly improves performance and scalability in a variety of domains. These include: single-task RL on Atari 2600 games with SoftMoEs, multi-task RL on Atari with large-scale ResNets, robotic manipulation with Q-transformers, playing Chess without search, and a language-agent Wordle task with high-capacity Transformers, achieving state-of-the-art results on these domains. Through careful analysis, we show that the benefits of categorical cross-entropy primarily stem from its ability to mitigate issues inherent to value-based RL, such as noisy targets and non-stationarity. Overall, we argue that a simple shift to training value functions with categorical cross-entropy can yield substantial improvements in the scalability of deep RL at little-to-no cost.

2024-07-07

Proceedings of the 41st International Conference on Machine Learning (publié)

doi.org

proceedings.mlr.press

Investigating Multi-Task Pretraining and Generalization in Reinforcement Learning

Bellemare Marc-Emmanuel

Google Brain

2023-01-31

ICLR.cc/2023/Conference (poster)

I NTRODUCING C OORDINATION IN C ONCURRENT R EIN - FORCEMENT L EARNING

Bellemare Marc-Emmanuel

Google Brain

Research on exploration in reinforcement learning has mostly focused on problems with a single agent interacting with an environment. Howeve… (voir plus)r many problems are better addressed by the concurrent reinforcement learning paradigm, where multiple agents operate in a common environment. Recent work has tackled the challenge of exploration in this particular setting (Dimakopoulou & Van Roy, 2018; Dimakopoulou et al., 2018). Nonetheless, they do not completely leverage the characteristics of this framework and agents end up behaving independently from each other. In this work we argue that coordination among concurrent agents is crucial for efficient exploration. We introduce coordination in Thompson Sampling based methods by drawing correlated samples from an agent’s posterior. We apply this idea to extend existing exploration schemes such as randomized least squares value iteration (RLSVI). Empirical results on simple toy tasks emphasize the merits of our approach and call attention to coordination as a key objective for efficient exploration in concurrent reinforcement learning.

2022-04-24

ICLR.cc/2022/Workshop/GMS (publié)

On Bonus-Based Exploration Methods in the Arcade Learning Environment

William Fedus

Marlos C. Machado

Bellemare Marc-Emmanuel

Research on exploration in reinforcement learning, as applied to Atari 2600 game-playing, has emphasized tackling difficult exploration prob… (voir plus)lems such as Montezuma's Revenge (Bellemare et al., 2016). Recently, bonus-based exploration methods, which explore by augmenting the environment reward, have reached above-human average performance on such domains. In this paper we reassess popular bonus-based exploration methods within a common evaluation framework. We combine Rainbow (Hessel et al., 2018) with different exploration bonuses and evaluate its performance on Montezuma's Revenge, Bellemare et al.'s set of hard of exploration games with sparse rewards, and the whole Atari 2600 suite. We find that while exploration bonuses lead to higher score on Montezuma's Revenge they do not provide meaningful gains over the simpler

2019-12-31

ICLR.cc/2020/Conference (poster)

doi.org

Benchmarking Bonus-Based Exploration Methods on the Arcade Learning Environment

William Fedus

Marlos C. Machado

Bellemare Marc-Emmanuel

This paper provides an empirical evaluation of recently developed exploration algorithms within the Arcade Learning Environment (ALE). We st… (voir plus)udy the use of different reward bonuses that incentives exploration in reinforcement learning. We do so by fixing the learning algorithm used and focusing only on the impact of the different exploration bonuses in the agent's performance. We use Rainbow, the state-of-the-art algorithm for value-based agents, and focus on some of the bonuses proposed in the last few years. We consider the impact these algorithms have on performance within the popular game Montezuma's Revenge which has gathered a lot of interest from the exploration community, across the the set of seven games identified by Bellemare et al. (2016) as challenging for exploration, and easier games where exploration is not an issue. We find that, in our setting, recently developed bonuses do not provide significantly improved performance on Montezuma's Revenge or hard exploration games. We also find that existing bonus-based methods may negatively impact performance on games in which exploration is not an issue and may even perform worse than

2019-08-05

ArXiv (prépublication)

arxiv.org

The Value Function Polytope in Reinforcement Learning

Robert Dadashi

Nicolas Roux

Dale Schuurmans

Bellemare Marc-Emmanuel

We establish geometric and topological properties of the space of value functions in finite state-action Markov decision processes. Our main… (voir plus) contribution is the characterization of the nature of its shape: a general polytope (Aigner et al., 2010). To demonstrate this result, we exhibit several properties of the structural relationship between policies and value functions including the line theorem, which shows that the value functions of policies constrained on all but one state describe a line segment. Finally, we use this novel perspective to introduce visualizations to enhance the understanding of the dynamics of reinforcement learning algorithms.

2019-05-23

Proceedings of the 36th International Conference on Machine Learning (publié)

proceedings.mlr.press

A Geometric Perspective on Optimal Representations for Reinforcement Learning

Bellemare Marc-Emmanuel

Will Dabney

Robert Dadashi

Pablo Samuel Castro

Nicolas Roux

Dale Schuurmans

Tor Lattimore

Clare Lyle

2018-12-31

NeurIPS (publié)

Sim-to-Real Transfer with Neural-Augmented Robot Simulation

Despite the recent successes of deep reinforcement learning, teaching complex motor skills to a physical robot remains a hard problem. While… (voir plus) learning directly on a real system is usually impractical, doing so in simulation has proven to be fast and safe. Nevertheless, because of the "reality gap," policies trained in simulation often perform poorly when deployed on a real system. In this work, we introduce a method for training a recurrent neural network on the differences between simulated and real robot trajectories and then using this model to augment the simulator. This Neural-Augmented Simulation (NAS) can be used to learn control policies that transfer significantly better to real environments than policies learned on existing simulators. We demonstrate the potential of our approach through a set of experiments on the Mujoco simulator with added backlash and the Poppy Ergo Jr robot. NAS allows us to learn policies that are competitive with ones that would have been learned directly on the real robot.

2018-10-22

Proceedings of The 2nd Conference on Robot Learning (publié)

proceedings.mlr.press

Approximate Exploration through State Abstraction

Bellemare Marc-Emmanuel

Although exploration in reinforcement learning is well understood from a theoretical point of view, provably correct methods remain impracti… (voir plus)cal. In this paper we study the interplay between exploration and approximation, what we call approximate exploration. Our main goal is to further our theoretical understanding of pseudo-count based exploration bonuses (Bellemare et al., 2016), a practical exploration scheme based on density modelling. As a warm-up, we quantify the performance of an exploration algorithm, MBIE-EB (Strehl and Littman, 2008), when explicitly combined with state aggregation. This allows us to confirm that, as might be expected, approximation allows the agent to trade off between learning speed and quality of the learned policy. Next, we show how a given density model can be related to an abstraction and that the corresponding pseudo-count bonus can act as a substitute in MBIE-EB combined with this abstraction, but may lead to either under- or over-exploration. Then, we show that a given density model also defines an implicit abstraction, and find a surprising mismatch between pseudo-counts derived either implicitly or explicitly. Finally we derive a new pseudo-count bonus alleviating this issue.

2018-08-28

ArXiv (prépublication)

arxiv.org

PixelVAE: A Latent Variable Model for Natural Images

Natural image modeling is a landmark challenge of unsupervised learning. Variational Autoencoders (VAEs) learn a useful latent representatio… (voir plus)n and model global structure well but have difficulty capturing small details. PixelCNN models details very well, but lacks a latent code and is difficult to scale for capturing large structures. We present PixelVAE, a VAE model with an autoregressive decoder based on PixelCNN. Our model requires very few expensive autoregressive layers compared to PixelCNN and learns latent codes that are more compressed than a standard VAE while still capturing most non-trivial structure. Finally, we extend our model to a hierarchy of latent variables at different scales. Our model achieves state-of-the-art performance on binarized MNIST, competitive performance on 64 × 64 ImageNet, and high-quality samples on the LSUN bedrooms dataset.

2016-12-31

ICLR (Poster) (publié)