Cohere's Multicultural Riddle Benchmark Enters Human Evaluation

Cohere Labs' multicultural riddle benchmark has entered the human evaluation phase, having already assessed over 61,000 model responses. The project is currently recruiting contributors to help advance its multilingual NLP evaluations.

2026-08-01 ~ 2026-08-01 · 2 related posts

• Cohere Launches Multilingual Model Human Evaluation, Seeks Contributors — Cohere_Labs · 2026-08-01
• Cohere's Multicultural Riddle Benchmark Enters Human Evaluation Phase — Cohere_Labs · 2026-08-01