Minimal Pair-Based Evaluation of Code-Switching

• URL: http://arxiv.org/abs/2506.01840v2
• Date: Fri, 25 Jul 2025 20:27:42 GMT
• Title: Minimal Pair-Based Evaluation of Code-Switching
• Authors: Igor Sterner, Simone Teufel,
• Abstract summary: Existing methods do not have wide language coverage, fail to account for the diverse range of code-switching phenomena, or do not scale.<n>We propose an intervention based on minimal pairs of CS. Each minimal pair contains one naturally occurring CS sentence and one minimally manipulated variant.<n>Our human experiments show that, for every language pair, bilinguals consistently prefer the naturally occurring CS sentence.
• Score: 2.100960337325026
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: There is a lack of an evaluation methodology that estimates the extent to which large language models (LLMs) use code-switching (CS) in the same way as bilinguals. Existing methods do not have wide language coverage, fail to account for the diverse range of CS phenomena, or do not scale. We propose an intervention based on minimal pairs of CS. Each minimal pair contains one naturally occurring CS sentence and one minimally manipulated variant. We collect up to 1,000 such pairs each for 11 language pairs. Our human experiments show that, for every language pair, bilinguals consistently prefer the naturally occurring CS sentence. Meanwhile our experiments with current LLMs show that the larger the model, the more consistently it assigns higher probability to the naturally occurring CS sentence than to the variant. In accordance with theoretical claims, the largest probability differences arise in those pairs where the manipulated material consisted of closed-class words.

Related papers

• Code-Switching and Syntax: A Large-Scale Experiment [2.100960337325026]
  We show that syntax alone is sufficient for an automatic system to distinguish between sentences in minimal pairs of code-switching humans.<n>The learnt syntactic patterns generalise well to unseen language pairs.
  arXiv  Detail & Related papers  (2025-06-02T16:32:14Z)
• ConCSE: Unified Contrastive Learning and Augmentation for Code-Switched Embeddings [4.68732641979009]
  This paper examines the Code-Switching (CS) phenomenon where two languages intertwine within a single utterance.<n>We highlight that the current Equivalence Constraint (EC) theory for CS in other languages may only partially capture English-Korean CS complexities.<n>We introduce a novel Koglish dataset tailored for English-Korean CS scenarios to mitigate such challenges.
  arXiv  Detail & Related papers  (2024-08-28T11:27:21Z)
• Understanding and Mitigating Language Confusion in LLMs [76.96033035093204]
  We evaluate 15 typologically diverse languages with existing and newly-created English and multilingual prompts.<n>We find that Llama Instruct and Mistral models exhibit high degrees of language confusion.<n>We find that language confusion can be partially mitigated via few-shot prompting, multilingual SFT and preference tuning.
  arXiv  Detail & Related papers  (2024-06-28T17:03:51Z)
• Roles of Scaling and Instruction Tuning in Language Perception: Model vs. Human Attention [58.817405319722596]
  This work compares the self-attention of several large language models (LLMs) in different sizes to assess the effect of scaling and instruction tuning on language perception. Results show that scaling enhances the human resemblance and improves the effective attention by reducing the trivial pattern reliance, while instruction tuning does not. We also find that current LLMs are consistently closer to non-native than native speakers in attention, suggesting a sub-optimal language perception of all models.
  arXiv  Detail & Related papers  (2023-10-29T17:16:40Z)
• Multimodal Chain-of-Thought Reasoning in Language Models [94.70184390935661]
  We propose Multimodal-CoT that incorporates language (text) and vision (images) modalities into a two-stage framework. Experimental results on ScienceQA and A-OKVQA benchmark datasets show the effectiveness of our proposed approach.
  arXiv  Detail & Related papers  (2023-02-02T07:51:19Z)
• A Multi-level Supervised Contrastive Learning Framework for Low-Resource Natural Language Inference [54.678516076366506]
  Natural Language Inference (NLI) is a growingly essential task in natural language understanding. Here we propose a multi-level supervised contrastive learning framework named MultiSCL for low-resource natural language inference.
  arXiv  Detail & Related papers  (2022-05-31T05:54:18Z)
• Call Larisa Ivanovna: Code-Switching Fools Multilingual NLU Models [1.827510863075184]
  Novel benchmarks for multilingual natural language understanding (NLU) include monolingual sentences in several languages, annotated with intents and slots. Existing benchmarks lack of code-switched utterances, which are difficult to gather and label due to complexity in the grammatical structure. Our work adopts recognized methods to generate plausible and naturally-sounding code-switched utterances and uses them to create a synthetic code-switched test set.
  arXiv  Detail & Related papers  (2021-09-29T11:15:00Z)
• Style Variation as a Vantage Point for Code-Switching [54.34370423151014]
  Code-Switching (CS) is a common phenomenon observed in several bilingual and multilingual communities. We present a novel vantage point of CS to be style variations between both the participating languages. We propose a two-stage generative adversarial training approach where the first stage generates competitive negative examples for CS and the second stage generates more realistic CS sentences.
  arXiv  Detail & Related papers  (2020-05-01T15:53:16Z)
• XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning [68.57658225995966]
  Cross-lingual Choice of Plausible Alternatives (XCOPA) is a typologically diverse multilingual dataset for causal commonsense reasoning in 11 languages. We evaluate a range of state-of-the-art models on this novel dataset, revealing that the performance of current methods falls short compared to translation-based transfer.
  arXiv  Detail & Related papers  (2020-05-01T12:22:33Z)
• Knowledge Distillation for Multilingual Unsupervised Neural Machine Translation [61.88012735215636]
  Unsupervised neural machine translation (UNMT) has recently achieved remarkable results for several language pairs. UNMT can only translate between a single language pair and cannot produce translation results for multiple language pairs at the same time. In this paper, we empirically introduce a simple method to translate between thirteen languages using a single encoder and a single decoder.
  arXiv  Detail & Related papers  (2020-04-21T17:26:16Z)

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.