TY - RPRT TI - Better Safe than Sorry: Evidence Accumulation Allows for Safe Reinforcement Learning AU - Akshat Agarwal AU - Abhinau Kumar V AU - Kyle Dunovan AU - Erik Peterson AU - Timothy Verstynen AU - Katia Sycara PY - 2018 UR - https://arxiv.org/abs/1809.09147 ID - 1809.09147 ER -