Saved in:
Bibliographic Details
Main Authors: Ye, Yangfan, Feng, Xiaocheng, Tang, Jialong, Cao, Xiayu, Zhang, Zihan, Feng, Xiachong, Yang, Baosong, Qin, Bing
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2606.01879
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910280197341184
author Ye, Yangfan
Feng, Xiaocheng
Tang, Jialong
Cao, Xiayu
Zhang, Zihan
Feng, Xiachong
Yang, Baosong
Qin, Bing
author_facet Ye, Yangfan
Feng, Xiaocheng
Tang, Jialong
Cao, Xiayu
Zhang, Zihan
Feng, Xiachong
Yang, Baosong
Qin, Bing
contents Existing research largely reduces cultural intelligence in LLMs to a knowledge-level problem, overlooking whether models can effectively utilize their acquired knowledge in realistic scenarios. To bridge this gap, we introduce CultureForest, a benchmark for \textit{Cultural Norm Grounded Reasoning}. Each question is grounded in a small set of atomic norms, enabling verifiable and attributable evaluation. CultureForest comprises 5,378 examples across 8 domains and 53 countries/regions, and supports a progressive evaluation from multiple-choice to open-ended generation. Extensive experiments reveal that even top-tier models degrade substantially in open-ended settings, accompanied by pronounced cross-region disparities. Through targeted analysis, we uncover several consistent patterns: (1) test-time reasoning yields limited gains and may exacerbate inequity; (2) models exhibit highly shared regional preference structures; (3) model responses are markedly conservative, especially under stricter cultural constraints; and (4) by disentangling cultural knowledge acquisition from cultural reasoning, we show that while LLMs possess substantial cultural knowledge, their performance is further bottlenecked by its effective use. These findings point to a necessary shift from knowledge-centric evaluation toward measuring knowledge-grounded reasoning.
format Preprint
id arxiv_https___arxiv_org_abs_2606_01879
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CultureForest: Understanding and Evaluating Cultural Norm Grounded Reasoning in LLMs
Ye, Yangfan
Feng, Xiaocheng
Tang, Jialong
Cao, Xiayu
Zhang, Zihan
Feng, Xiachong
Yang, Baosong
Qin, Bing
Computation and Language
Existing research largely reduces cultural intelligence in LLMs to a knowledge-level problem, overlooking whether models can effectively utilize their acquired knowledge in realistic scenarios. To bridge this gap, we introduce CultureForest, a benchmark for \textit{Cultural Norm Grounded Reasoning}. Each question is grounded in a small set of atomic norms, enabling verifiable and attributable evaluation. CultureForest comprises 5,378 examples across 8 domains and 53 countries/regions, and supports a progressive evaluation from multiple-choice to open-ended generation. Extensive experiments reveal that even top-tier models degrade substantially in open-ended settings, accompanied by pronounced cross-region disparities. Through targeted analysis, we uncover several consistent patterns: (1) test-time reasoning yields limited gains and may exacerbate inequity; (2) models exhibit highly shared regional preference structures; (3) model responses are markedly conservative, especially under stricter cultural constraints; and (4) by disentangling cultural knowledge acquisition from cultural reasoning, we show that while LLMs possess substantial cultural knowledge, their performance is further bottlenecked by its effective use. These findings point to a necessary shift from knowledge-centric evaluation toward measuring knowledge-grounded reasoning.
title CultureForest: Understanding and Evaluating Cultural Norm Grounded Reasoning in LLMs
topic Computation and Language
url https://arxiv.org/abs/2606.01879