Why Should I Trust You, Bellman? The Bellman Error is a Poor Replacement for Value Error
Ok i thought this paper had big implications, but it feels more minor after reading it. The bellman equation is a surrogate of how much the value function is actually off (this is measured via the full return ).
Citation
(Fujimoto et al., 2022)
Fujimoto, S., Meger, D., Precup, D., Nachum, O., & Gu, S. S. (2022). Why Should I Trust You, Bellman? The Bellman Error is a Poor Replacement for Value Error. https://arxiv.org/abs/2201.12417