Your Algorithm: Giving People More Ways to Personalize Facebook Recommendations

Why it matters

Facebook has published a technical account of Your algorithm, a suite of AI-powered tools that lets people on Facebook shape their Feed and Reels recommendations by stating preferences in their own words. The system grew out of a limitation of traditional recommender systems: they learn from implicit engagement signals, which scale well but adapt slowly and capture little nuance. Your algorithm addresses this with a collection of large language models that understand, store, and track stated preferences alongside user behavior. The interface shows a summary of recent viewing activity plus editable lists of interests the user wants more or less of. The engineering account explains how each preference moves through a full lifecycle: profile generation, preference management, ranking, and delivery optimization. New interests trigger dedicated retrieval sources so relevant content actually enters the ranking pipeline, while a control layer keeps any single interest from dominating the feed. Suppressed topics are enforced through layered filtering across ranking, reranking, and the device, with cache invalidation and supplementary fetches ensuring changes take effect quickly.