TY  - RPRT
TI  - Reward Model Ensembles Help Mitigate Overoptimization
AU  - Thomas Coste
AU  - Usman Anwar
AU  - Robert Kirk
AU  - David Krueger
PY  - 2024
UR  - https://arxiv.org/abs/2310.02743
ID  - 2310.02743
ER  -