Scaling Reranked Retrieval report: some rerankers degrade when given too many documents

CShorten30 · x · 2026-11-18

CShorten30's team released the "Scaling Reranked Retrieval" research report; amplifier mrdrozdov highlighted two key findings:

Chart data: on five BRIGHT subsets, hybrid retrieval top-20 without reranking yields Recall@20 of just 0.282; ZeroEntropy zerank-2 with a k=2,000 pool raises it to 0.560; adding a Listwise GPT-5.6 Luna stage reorders the same 20 docs (Recall@20 pinned) and pushes Success@1 to 0.478.

Related event: Weaviate Releases "Scaling Reranked Retrieval" Research with Open-Source Code(10 posts)→

Original post →

More from Research

Research channel →