From the source
Lead story
Ask your AI
Top stories
Models & availability
Latest
Perplexity· PerplexityQ2D-Web: Evaluating First-Stage Retrievers at Scale
A benchmark for evaluating retrieval in agentic RAG systems at web scale.
Perplexity introduced Q2D-Web, a private benchmark and public leaderboard for evaluating first-stage retrieval in agentic RAG systems.
The benchmark includes 190 million web documents and 69,721 agent-reformulated queries across ten languages, with three relevance-judgment sets to reduce bias and false negatives.
From the source
Q2D-Web is built to evaluate embedding models on large-scale web search. It consists of 190 million web documents and 69,721 agent-reformulated queries in ten languages, sampled over nine months of PII-free production search traffic.