Joshua Romoff Peter Henderson 0002 David Kanaa Emmanuel Bengio Ahmed Touati Pierre-Luc Bacon Joelle Pineau TDprop: Does Jacobi Preconditioning Help Temporal Difference Learning? 2020 abs/2007.02786 CoRR https://arxiv.org/abs/2007.02786 db/journals/corr/corr2007.html#abs-2007-02786