TY  - RPRT
TI  - Can We Really Learn One Representation to Optimize All Rewards?
AU  - Chongyi Zheng
AU  - Royina Karegoudra Jayanth
AU  - Benjamin Eysenbach
PY  - 2026
UR  - https://arxiv.org/abs/2602.11399
ID  - 2602.11399
ER  -