TY - RPRT TI - Towards More Efficient, Robust, Instance-adaptive, and Generalizable Sequential Decision making AU - Zhiyong Wang PY - 2025 UR - https://arxiv.org/abs/2504.09192 ID - 2504.09192 ER -