LinkedIn Details Its Profile Relevance Model for Surfacing Posts From Accounts Users Don't Follow
TLDR
LinkedIn's Profile Relevance model uses LLM-generated embeddings to show users more relevant posts from people they don't follow.
Optimixed Overview
Optimixed's original report on this story.
LinkedIn's Profile Relevance model uses LLM-generated embeddings to surface posts from people users don't follow, based on relevance to each user. Social Media Today covers the company's explanation of the update. The article text is brief, so details beyond this description aren't available.
How we got here: earlier coverage
- Oct 13Meta adds LLM-based ranking to Facebook feed, accepts typed interest requests within sessions
Meta is adding LLM-based ranking to Facebook's feed and Reels, allowing users to type interest requests that update dynamically within a session, though the exact signal creators and advertisers receive remains unclear.
- Aug 17New research shows individual LinkedIn profiles generate 75% of AI citations, outpacing company pages
Research from Meltwater and LinkedIn analyzing 9.5 million citations found that individual profiles drive three-quarters of all LinkedIn citations in AI systems.
- Aug 11LinkedIn Launches AI Playbook to Help Professionals Boost Profile Discovery and Credibility
LinkedIn has released a new playbook focused on using AI to optimize professional profiles for better discoverability and credibility.
- Aug 10LinkedIn Updates Feed Display to Drive More Post Replies and Comments
LinkedIn is adjusting its feed algorithm to prioritize relevance and rank interactions based on engagement quality rather than volume.