SCALING UP RL IS ALL THE RAGE RIGHT NOW

Summary from the TLDR AI newsletter

Reinforcement learning (RL) will continue to lead to more gains because when done well, it is a lot more leveraged, more responsive to feedback, and superior to supervised fine-tuning. Researchers are likely to discover more about RL as rollout lengths continue to expand. There are a lot more S curves to discover, possibly specific to large language models and without analogues in game/robotics-like environments.