Meta Introduces New Algorithm Personalization Tools for Facebook Feeds
The company is utilizing large language models to improve algorithm personalization across its platform.
Managing Recommendations
Meta is rolling out a new algorithm personalization feature that aims to make it easier for users to explain what they do and do not want to see in their recommendations.
With the update, users can access specific feed posts and Reels to open a new menu. From there, specific topics can be added to personal interests or to a not interested list. There is also a text box where Meta prompts users to describe in their own words what they want to see in their feed.
Large Language Models
The update mirrors a similar change introduced on Instagram. Meta states that this functionality is possible because the platform now uses large language models in its ranking systems, moving beyond simple engagement inference to incorporate the context and intent behind natural language.
Previous iterations of similar features on other apps have shown certain limitations regarding requests for fewer advertisements or specific content types, but the new tool marks a shift toward more direct user input in social media curation.
