You can now decide what appears on your Facebook Feed
The changes are supported by large language models (LLMs), which allow Facebook's recommendation systems to interpret the context, interests, and intent behind users' written feedback.
-
Courtesy of Meta
Announced by Meta on October 13, the feature, called “Your Algorithm,” introduces a more interactive approach to content recommendations, allowing users to tell Facebook what they want to see more or less of without relying solely on their previous online activities.
The tool provides a dedicated space where users can add and manage topics based on their interests, giving them greater control over the posts and videos recommended by the platform.
Accessible through the three-dot menu on Facebook posts, the feature allows users to communicate their preferences using natural language rather than depending entirely on existing engagement signals.
Meta said the updated recommendation system can respond to user preferences in real time, adjusting the content displayed in Feed and Reels within the same browsing session.
The company is also introducing additional feedback options that allow users to explain why they are interested or uninterested in particular posts.
After selecting “Interested” or “Not interested,” users may receive follow-up questions asking them to provide more details about the type of content they want to see.
These responses can be written in their own words, allowing Facebook to better understand individual preferences and refine its recommendations accordingly.
The changes are supported by large language models (LLMs), which allow Facebook's recommendation systems to interpret the context, interests, and intent behind users' written feedback.
Previously, Facebook's ranking systems primarily relied on behavioral signals, including likes, comments, video views, and other interactions, to determine which content would appear in users' feeds.
With the integration of LLMs, the platform can combine these existing signals with direct feedback to generate more personalized recommendations.
According to Meta, the updated system can interpret more detailed preferences, allowing users to communicate changes in their interests as they browse the platform.
The company has also redesigned the three-dot menu on posts to make the new customization and feedback tools easier to access.
The update forms part of Meta's broader efforts to introduce AI-powered personalization tools across its social media platforms.
Meta did not specify when the feature would become available to all users, including those in the Philippines.
Key points
- The changes are supported by large language models (LLMs), which allow Facebook's recommendation systems to interpret the context, interests, and intent behind users' written feedback.