Engineering The Next Generation Of Linkedins Feed
- Source
- linkedin.com
- Author
- LinkedIn Engineering
- Date
Context
LinkedIn's March 2026 feed redesign uses the order of a person's interactions to help decide which posts to show. Its engineering account describes a ranking model that processes a sequence of posts together with actions such as likes, comments and long reading time. This lets recent activity influence how earlier interests are interpreted, rather than treating each encounter independently.
A separate retrieval stage first finds candidate posts by comparing , numerical representations that a fine-tuned produces for members and for posts. Ranking then orders those candidates using interaction history. Together, these mechanisms are intended to surface relevant professional material as interests change, including posts from people the member does not follow. The article presents the company's system design and intended experience, not an independent measurement of member benefit.
- embedding — A list of numbers representing a piece of text's meaning, so that similar meanings end up numerically close and can be searched.
- LLM — A large language model — the neural network behind tools like Claude and ChatGPT, trained on huge amounts of text to predict what comes next.
- transformer — The neural network architecture behind modern AI models, built on attention — letting every word directly consider every other word in parallel.
- articleBuilding The Next Generation Of Job Search At LinkedinLinkedIn Engineering
- articleRebuilding Linkedins Follows Recommendations With Llm Based Semantic Retrieval And RankingLinkedIn Engineering
- articleFaster Than Light Optimizing Generative Recommender Training Efficiency At LinkedinLinkedIn Engineering
Checking sign-in…
Loading comments…