CQ-Bench Paper Accepted by TMLR: Evaluating LLM Cultural Intelligence

jieyuzhao11 · x · 2026-09-01

A paper introducing CQ-Bench has been accepted to TMLR. CQ-Bench is a benchmark designed to evaluate the cultural intelligence of LLMs through dialogue, testing whether models can understand implicit cultural values in conversation as naturally as humans do.

Original post →

More from Research

Research channel →