two-tower
2 free lessons tagged two-tower across AI. Each one is a short sequence of focused steps with narration and a five-question quiz at the end — take them in any order, no signup required.
Two Towers: How Taste Becomes Geometry
Candidate generation's workhorse is the two-tower model: one network embeds users, another embeds items, and relevance becomes a dot product in shared space. This lesson traces the idea from collaborative filtering through matrix factorization to trained towers, why the towers must never mix features, and the cold-start problem geometry alone cannot solve.
The two-stage funnel: retrieval then ranking
Every feed has the same impossible job: pick ten items out of millions, in under a tenth of a second. The answer is a funnel. Learn why recommenders split into a cheap retrieval stage and an expensive ranking stage, how two-tower models make retrieval possible, and why nearest-neighbour search is the trick that makes the whole thing fit in a budget.