While training a language model using reinforcement learning from human feedback (RLHF), reward models are typically tuned to ...

The objective is to: learn to make polite suggestions and requests ... To become familiar with differences between spoken and written language. To read short newspaper articles and other non-fiction, ...