TY  - RPRT
TI  - Reinforcement Learning from Human Feedback
AU  - Nathan Lambert
PY  - 2026
UR  - https://arxiv.org/abs/2504.12501
ID  - 2504.12501
ER  -