paper-with-me

홈 › Papers

In Support of Over-Parametrization in Deep Reinforcement Learning: an Empirical Study

2019-05-17 · ICML Workshop Deep_Phenomen 2019 6 · Brady Neal, Ioannis Mitliagkas

There is significant recent evidence in supervised learning that, in the over-parametrized setting, wider networks achieve better test error. In other words, the bias-variance tradeoff is not directly observable when increasing network width arbitrarily. We investigate whether a corresponding phenomenon is present in reinforcement learning. We experiment on four OpenAI Gym environments, increasing the width of the value and policy networks beyond their prescribed values. Our empirical results lend support to this hypothesis. However, tuning the hyperparameters of each network width separately remains as important future work in environments/algorithms where the optimal hyperparameters vary noticably across widths, confounding the results when the same hyperparameters are used for all widths.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

Deep Reinforcement LearningOpenAI Gymreinforcement-learningReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Overparametrization improves robustness against adversarial attacks: A replication study

2022-02-20 · Ali Borji

Overparametrization has become a de facto standard in machine learning. Despite numerous efforts, our understanding of how and where overparametrization helps model accuracy and robustness is still limited. To this end, …

An Empirical Analysis of Measure-Valued Derivatives for Policy Gradients

2021-07-20 · João Carvalho, Davide Tateo, Fabio Muratore, Jan Peters

Reinforcement learning methods for robotics are increasingly successful due to the constant development of better policy gradient techniques. A precise (low variance) and accurate (low bias) gradient estimator is crucial…

A Proof of Learning Rate Transfer under $μ$P

2025-11-03 · Soufiane Hayou arxiv

We provide the first proof of learning rate transfer with width in a linear multi-layer perceptron (MLP) parametrized with $μ$P, a neural network parameterization designed to ``maximize'' feature learning in the infinite…

A posteriori learning for quasi-geostrophic turbulence parametrization

2022-04-08 · Hugo Frezat, Julien Le Sommer, Ronan Fablet, Guillaume Balarac 외

The use of machine learning to build subgrid parametrizations for climate models is receiving growing attention. State-of-the-art strategies address the problem as a supervised learning task and optimize algorithms that …

On the Global Optimum Convergence of Momentum-based Policy Gradient

2021-10-19 · Yuhao Ding, Junzi Zhang, Javad Lavaei

Policy gradient (PG) methods are popular and efficient for large-scale reinforcement learning due to their relative stability and incremental nature. In recent years, the empirical success of PG methods has led to the de…