Saved in:
Bibliographic Details
Main Authors: Yang, Qingyan, Wang, Tongxi, Luo, Yunsheng
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.16217
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911393251328000
author Yang, Qingyan
Wang, Tongxi
Luo, Yunsheng
author_facet Yang, Qingyan
Wang, Tongxi
Luo, Yunsheng
contents Code-mixing is increasingly prevalent in interactions between humans and large language models, yet existing work often reduces it to a translation or convertibility problem, making it difficult to assess whether a model's switching behavior is context-appropriate and aligned with human conventions. We introduce ChiEngMixBench, the first benchmark designed to evaluate code-mixing ability in authentic community contexts, built upon a general construction pipeline that enables scalable dataset development across domains and bilingual pairs. ChiEngMixBench formulates code-mixing as a cognitive alignment problem, characterized by two complementary signals: Spontaneity and Naturalness. Empirical evaluation shows that our metrics can systematically distinguish code-mixing performance across models. Beyond benchmarking, we further uncover an implicitly emergent Terminology Layering Strategy, a phenomenon consistent with the Matrix Language Frame (MLF) theory, indicating structured cognitive alignment between multilingual large language models and human communication.
format Preprint
id arxiv_https___arxiv_org_abs_2601_16217
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle ChiEngMixBench: Evaluating Large Language Models on Spontaneous and Natural Chinese-English Code-Mixed Generation
Yang, Qingyan
Wang, Tongxi
Luo, Yunsheng
Computation and Language
Artificial Intelligence
Code-mixing is increasingly prevalent in interactions between humans and large language models, yet existing work often reduces it to a translation or convertibility problem, making it difficult to assess whether a model's switching behavior is context-appropriate and aligned with human conventions. We introduce ChiEngMixBench, the first benchmark designed to evaluate code-mixing ability in authentic community contexts, built upon a general construction pipeline that enables scalable dataset development across domains and bilingual pairs. ChiEngMixBench formulates code-mixing as a cognitive alignment problem, characterized by two complementary signals: Spontaneity and Naturalness. Empirical evaluation shows that our metrics can systematically distinguish code-mixing performance across models. Beyond benchmarking, we further uncover an implicitly emergent Terminology Layering Strategy, a phenomenon consistent with the Matrix Language Frame (MLF) theory, indicating structured cognitive alignment between multilingual large language models and human communication.
title ChiEngMixBench: Evaluating Large Language Models on Spontaneous and Natural Chinese-English Code-Mixed Generation
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2601.16217