TY  - RPRT
TI  - Modular Multitask Reinforcement Learning with Policy Sketches
AU  - Jacob Andreas
AU  - Dan Klein
AU  - Sergey Levine
PY  - 2017
UR  - https://arxiv.org/abs/1611.01796
ID  - 1611.01796
ER  -