We’re exploring using LLMs to deal with these challenges. Our giant language fashions like GPT-4 can perceive and generate pure language, making them relevant to content material moderation. The fashions could make moderation judgments based mostly on coverage tips supplied to them.
With this technique, the method of growing and customizing content material insurance policies is trimmed down from months to hours.Â
- As soon as a coverage guideline is written, coverage consultants can create a golden set of information by figuring out a small variety of examples and assigning them labels based on the coverage. Â
- Then, GPT-4 reads the coverage and assigns labels to the identical dataset, with out seeing the solutions.Â
- By analyzing the discrepancies between GPT-4’s judgments and people of a human, the coverage consultants can ask GPT-4 to provide you with reasoning behind its labels, analyze the paradox in coverage definitions, resolve confusion and supply additional clarification within the coverage accordingly. We are able to repeat steps 2 and three till we’re happy with the coverage high quality.
This iterative course of yields refined content material insurance policies which might be translated into classifiers, enabling the deployment of the coverage and content material moderation at scale.
Optionally, to deal with giant quantities of information at scale, we are able to use GPT-4’s predictions to fine-tune a a lot smaller mannequin.
