Meta went in the opposite direction to LinkedIn: it keeps an ecosystem of more than a thousand specialized models arranged as a multi-stage retrieval funnel, ending in a multi-objective value model that extends the objective past raw engagement to avoid-content signals.

Meta arranges Instagram's recommendation system as a funnel of over a thousand specialized models across four stages (retrieval, early-stage two-tower ranking, heavy late-stage ranking, and a final diversity/integrity pass). Its value model extends beyond raw engagement to weight saves positively and subtract signals like 'See Fewer Posts Like This,' because many competing objectives are easier to tune and audit as separate stages. ✦ AI generated

Author (article How to Fight Clickbait: Meta, LinkedIn & YouTube Case Studies) · ByteByteGo Newsletter · 2026-08-10 · original ↗

Meta arranges Instagram's recommendation system as a multi-stage funnel, consisting of an ecosystem of more than a thousand models supporting it. Candidates pass through a sequence of stages, and each stage applies a more expensive model to a smaller set of surviving candidates... In Meta's approach, the late-stage model predicts many possible user actions at once, and a value model combines those predictions into a single score. That combination adds weight for positive actions, such as a likely save, and subtracts weight for predicted negative actions, such as a "See Fewer Posts Like This" tap. The basic objective of this extends past raw engagement to include signals about content a user would prefer to avoid... Many competing objectives, including engagement, diversity, integrity, and creator fairness, are easier to tune and audit as separate stages than as one model. Of course, the cost to this is operational complexity, which is the exact complexity LinkedIn set out to reduce.

Read full article ↗excerpt · fair-use quotation