Excited to see them almost certainly combine their RL expertise with their LLM expertise to encourage reasoning. It’s been the most obvious thing since the invention of LLMs, and I’m sure they will figure it out or deepmind will. We all know its coming. Excited for the near future.
Excited to see them almost certainly combine their RL expertise with their LLM expertise to encourage reasoning. It’s been the most obvious thing since the invention of LLMs, and I’m sure they will figure it out or deepmind will. We all know its coming. Excited for the near future.