CodeMixBench: Evaluating Code-Mixing Capabilities of LLMs Across 18 Languages

• URL: http://arxiv.org/abs/2507.18791v2
• Date: Sun, 07 Sep 2025 11:57:23 GMT
• Title: CodeMixBench: Evaluating Code-Mixing Capabilities of LLMs Across 18 Languages
• Authors: Yilun Yang, Yekun Chai,
• Abstract summary: Code-mixing, the practice of switching languages within a conversation, poses unique challenges for traditional NLP.<n>Existing benchmarks are limited by their narrow language pairs and tasks, failing to adequately assess large language models' (LLMs) code-mixing abilities.<n>We introduce CodeMixBench, a comprehensive benchmark covering eight tasks, including three specific to LLMs and five traditional NLP tasks, and 18 languages across seven language families.
• Score: 10.15537631183956
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Code-mixing, the practice of switching between languages within a conversation, poses unique challenges for traditional NLP. Existing benchmarks are limited by their narrow language pairs and tasks, failing to adequately assess large language models' (LLMs) code-mixing abilities. Despite the recognized importance of code-mixing for multilingual users, research on LLMs in this context remains sparse. Additionally, current techniques for synthesizing code-mixed data are underdeveloped to generate code-mixing. In response, we introduce CodeMixBench, a comprehensive benchmark covering eight tasks, including three specific to LLMs and five traditional NLP tasks, and 18 languages across seven language families. We also propose a new method for generating large-scale synthetic code-mixed texts by combining word substitution with GPT-4 prompting. Our evaluation reveals consistent underperformance of LLMs on code-mixed datasets involving different language families. Enhancements in training data size, model scale, and few-shot learning could improve their performance. The code and dataset are available at https://github.com/Jeromeyluck/CodeMixBench.

Related papers

• Can Large Language Models Understand, Reason About, and Generate Code-Switched Text? [26.210664542372168]
  Code-switching is a pervasive phenomenon in multilingual communication, yet the robustness of large language models (LLMs) in mixed-language settings remains insufficiently understood.<n>We introduce CodeMixQA, a novel benchmark with high-quality human annotations, comprising 16 diverse parallel code-switched language-pair variants.<n>We analyze the reasoning behavior of LLMs on code-switched question-answering tasks, shedding light on how models process and reason over mixed-language inputs.
  arXiv  Detail & Related papers  (2026-01-12T02:52:38Z)
• IFEvalCode: Controlled Code Generation [69.28317223249358]
  The paper introduces forward and backward constraints generation to improve the instruction-following capabilities of Code LLMs.<n>The authors present IFEvalCode, a multilingual benchmark comprising 1.6K test samples across seven programming languages.
  arXiv  Detail & Related papers  (2025-07-30T08:08:48Z)
• Can LLMs Replace Humans During Code Chunking? [2.4056836012742]
  Large language models (LLMs) have become essential tools in computer science, especially for tasks involving code understanding and generation.<n>This paper examines the application of LLMs in the modernization of legacy government code written in ALC and MUMPS.
  arXiv  Detail & Related papers  (2025-06-24T13:02:35Z)
• CodeMixBench: Evaluating Large Language Models on Code Generation with Code-Mixed Prompts [0.0]
  We introduce CodeMixBench, a novel benchmark to evaluate robustness of large language models (LLMs) on code generation from code-mixed prompts.<n>We comprehensively evaluate a diverse set of open-source code generation models ranging from 1.5B to 15B parameters.<n>Our results show that code-mixed prompts consistently degrade Pass@1 performance compared to their English-only counterparts.
  arXiv  Detail & Related papers  (2025-05-08T08:55:32Z)
• CHAI for LLMs: Improving Code-Mixed Translation in Large Language Models through Reinforcement Learning with AI Feedback [11.223762031003671]
  Large Language Models (LLMs) have demonstrated remarkable capabilities across various NLP tasks but struggle with code-mixed (or code-switched) language understanding.<n>This paper proposes CHAI, a novel framework for improving the ability of multilingual LLMs to handle code-mixed languages.<n>Our analysis shows that CHAI-powered LLMs outperform state-of-the-art open-source LLMs by 25.66% (in terms of win rate adjudicated by human annotators) in code-mixed translation tasks.
  arXiv  Detail & Related papers  (2024-11-13T22:56:00Z)
• Crystal: Illuminating LLM Abilities on Language and Code [58.5467653736537]
  We propose a pretraining strategy to enhance the integration of natural language and coding capabilities. The resulting model, Crystal, demonstrates remarkable capabilities in both domains.
  arXiv  Detail & Related papers  (2024-11-06T10:28:46Z)
• Linguistics Theory Meets LLM: Code-Switched Text Generation via Equivalence Constrained Large Language Models [16.82812708514889]
  Code-switching, the phenomenon of alternating between two or more languages in a single conversation, presents unique challenges for Natural Language Processing (NLP) Most existing research focuses on either syntactic constraints or neural generation, with few efforts to integrate linguistic theory with large language models (LLMs) for generating natural code-switched text. We introduce EZSwitch, a novel framework that combines Equivalence Constraint Theory (ECT) with LLMs to produce linguistically valid and fluent code-switched text.
  arXiv  Detail & Related papers  (2024-10-30T03:03:32Z)
• Think Carefully and Check Again! Meta-Generation Unlocking LLMs for Low-Resource Cross-Lingual Summarization [108.6908427615402]
  Cross-lingual summarization ( CLS) aims to generate a summary for the source text in a different target language.<n>Currently, instruction-tuned large language models (LLMs) excel at various English tasks.<n>Recent studies have shown that LLMs' performance on CLS tasks remains unsatisfactory even with few-shot settings.
  arXiv  Detail & Related papers  (2024-10-26T00:39:44Z)
• Code-Mixer Ya Nahi: Novel Approaches to Measuring Multilingual LLMs' Code-Mixing Capabilities [3.359458926468223]
  Rule-Based Prompting is a novel prompting technique to generate code-mixed sentences. We measure and compare the code-mixed MT abilities of 3 popular multilingual LLMs. We also use $k$-shot prompting to gauge the code-mixed to English translation abilities of multilingual LLMs.
  arXiv  Detail & Related papers  (2024-10-14T20:40:36Z)
• Crosslingual Capabilities and Knowledge Barriers in Multilingual Large Language Models [62.91524967852552]
  Large language models (LLMs) are typically multilingual due to pretraining on diverse multilingual corpora.<n>But can these models relate corresponding concepts across languages, i.e., be crosslingual?<n>This study evaluates state-of-the-art LLMs on inherently crosslingual tasks.
  arXiv  Detail & Related papers  (2024-06-23T15:15:17Z)
• IRCoder: Intermediate Representations Make Language Models Robust Multilingual Code Generators [49.903001442804594]
  This work investigates the prospect of leveraging compiler intermediate representations (IR) to improve the multilingual capabilities of Code-LMs. We first compile SLTrans, a parallel dataset consisting of nearly 4M self-contained source code files. Next, we carry out continued causal language modelling training on SLTrans, forcing the Code-LMs to learn the IR language. Our resulting models, dubbed IRCoder, display sizeable and consistent gains across a wide variety of code generation tasks and metrics.
  arXiv  Detail & Related papers  (2024-03-06T17:52:08Z)
• Code Needs Comments: Enhancing Code LLMs with Comment Augmentation [91.52444946362547]
  We introduce a novel data augmentation method that generates comments for existing code, coupled with a data filtering strategy that filters out code data poorly correlated with natural language. We conducted experiments on three code-focused Large Language Models and observed consistent improvements in performance on two widely-used programming skill benchmarks.
  arXiv  Detail & Related papers  (2024-02-20T13:56:38Z)
• CodeFuse-13B: A Pretrained Multi-lingual Code Large Language Model [58.127534002232096]
  This paper introduces CodeFuse-13B, an open-sourced pre-trained code LLM. It is specifically designed for code-related tasks with both English and Chinese prompts. CodeFuse achieves its effectiveness by utilizing a high quality pre-training dataset.
  arXiv  Detail & Related papers  (2023-10-10T02:38:44Z)
• Prompting Multilingual Large Language Models to Generate Code-Mixed Texts: The Case of South East Asian Languages [47.78634360870564]
  We explore prompting multilingual models to generate code-mixed data for seven languages in South East Asia (SEA) We find that publicly available multilingual instruction-tuned models such as BLOOMZ are incapable of producing texts with phrases or clauses from different languages. ChatGPT exhibits inconsistent capabilities in generating code-mixed texts, wherein its performance varies depending on the prompt template and language pairing.
  arXiv  Detail & Related papers  (2023-03-23T18:16:30Z)
• Massively Multilingual Shallow Fusion with Large Language Models [62.76735265311028]
  We train a single multilingual language model (LM) for shallow fusion in multiple languages. Compared to a dense LM of similar computation during inference, GLaM reduces the WER of an English long-tail test set by 4.4% relative. In a multilingual shallow fusion task, GLaM improves 41 out of 50 languages with an average relative WER reduction of 3.85%, and a maximum reduction of 10%.
  arXiv  Detail & Related papers  (2023-02-17T14:46:38Z)
• LEVER: Learning to Verify Language-to-Code Generation with Execution [64.36459105535]
  We propose LEVER, a simple approach to improve language-to-code generation by learning to verify the generated programs with their execution results. Specifically, we train verifiers to determine whether a program sampled from the LLMs is correct or not based on the natural language input, the program itself and its execution results. LEVER consistently improves over the base code LLMs(4.6% to 10.9% with code-davinci) and achieves new state-of-the-art results on all of them.
  arXiv  Detail & Related papers  (2023-02-16T18:23:22Z)
• MIPE: A Metric Independent Pipeline for Effective Code-Mixed NLG Evaluation [1.2559148369195197]
  Code-mixing is a phenomenon of mixing words and phrases from two or more languages in a single utterance of speech and text. Various widely popular metrics perform poorly with the code-mixed NLG tasks. We present a metric independent evaluation pipeline MIPE that significantly improves the correlation between evaluation metrics and human judgments.
  arXiv  Detail & Related papers  (2021-07-24T05:24:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.