Deep Reinforcement Learning from Human Preferences
A foundational 2017 paper that proposed a method for training agents via deep reinforcement learning guided by human preference comparisons rather than explicit reward functions. It constitutes one of the core theoretical bases of the RLHF technical approach.
Timeline (last 90 days)
In 2017, Christiano and colleagues published the paper "Deep Reinforcement Learning from Human Preferences," laying the theoretical foundation for the RLHF direction.
Christiano published the paper "Deep Reinforcement Learning from Human Preferences" in 2017, which is considered one of the foundational works of the RLHF technical approach.
All Facts (2)
In 2017, Christiano and colleagues published the paper "Deep Reinforcement Learning from Human Preferences," laying the theoretical foundation for the RLHF direction.
70%UnverifiedChristiano published the paper "Deep Reinforcement Learning from Human Preferences" in 2017, which is considered one of the foundational works of the RLHF technical approach.
50%