Cohere responds to benchmark criticism: still leads EmbeddingGemma 2 on standard nDCG@10
cohere · x · 2026-10-18
Responding to questions about whether RCP-nDCG is its own methodology, Cohere confirms it is, but says it still leads Google's EmbeddingGemma 2 on standard nDCG@10 as well: ViDoRe v3 Regular text-search average of 59.3 vs 47.6.
This is a follow-up to Cohere's earlier claim that Embed 5 Fast leads on over 95% of 163 datasets.
Related event: Cohere Claims Embed 5 Fast Beats Google's Embedding Model Across Benchmarks(2 posts)→
More from Models
- Perplexity's decision model wins Catan tournament at under half a cent per game — denisyarats · 2026-10-18
- DeepSeek V4.1 Flash 1017 aces 10-task coding benchmark at under 5 cents per task — Fabulous_Pollution10 · 2026-10-18
- Scaling01 Slams DeepSeek's New Model as 'Can't Ship Good Models Anymore', Drama Ensues — teortaxesTex · 2026-10-18
- From 4K to 1M tokens: context windows grew 250x in four years since GPT-3.5 — khademinori · 2026-10-18
- Ollama rolls out updated model to DeepSeek-V4.1-Flash users — ollama · 2026-10-18
- DeepSeek models reason about the grader on LisanBench, suggesting RL on the benchmark — scaling01 · 2026-10-18