This is a paper that was referenced when talking about how certain architectures are better for certain problems.

Citation

(Furuta et al., 2021)

Furuta, H., Kozuno, T., Matsushima, T., Matsuo, Y., & Gu, S. S. (2021). Co-Adaptation of Algorithmic and Implementational Innovations in Inference-based Deep Reinforcement Learning. https://arxiv.org/pdf/2103.17258