TY - RPRT TI - Towards a Practical Understanding of Lagrangian Methods in Safe Reinforcement Learning AU - Lindsay Spoor AU - Álvaro Serra-Gómez AU - Aske Plaat AU - Thomas Moerland PY - 2026 UR - https://arxiv.org/abs/2510.17564 ID - 2510.17564 ER -