OpenAI details CriticGPT, a model based on GPT-4 to catch errors in ChatGPT's code output, assisting human trainers in assessing and spotting errors (Will Knight/Wired)

https://ift.tt/hdQTHFj

Will Knight / Wired:
OpenAI details CriticGPT, a model based on GPT-4 to catch errors in ChatGPT's code output, assisting human trainers in assessing and spotting errors  —  Having humans rate a language model's outputs produced clever chatbots.  OpenAI says adding AI to the loop could help make them even smarter and more reliable.

from Techmeme https://ift.tt/ZeOD7TB