← back to paper
arxiv: 2604.17551 · 2 revisions
SVL: Goal-Conditioned Reinforcement Learning as Survival Learning