Mar. - Apr. 2025

I do not have an intuitive understanding of the successor representation. I opened Barto & Sutton (2018) in mid 2022, but then I shut it because I didn't like some of the early assumptions. Entering graduate school, I would like to be able to speak the language of the reinforcement learning community. I propose spending these two months getting into the details and, tentatively, I plan to culminate this period with a recitation from Carvalho et al.'s recent paper.