# Semantic Label Drift in Cross-Cultural Translation

arXiv: https://arxiv.org/abs/2510.25967 · HTML: https://mlsanity.com/paper/2510.25967/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 44 references that are in the map.
- First posted on arXiv 29 Oct 2025.

## Builds on (most cited first, 12 of 12)

- [BERTopic: Neural topic modeling with a class-based TF-IDF procedure](https://mlsanity.com/paper/2203.05794/) (2022)
- [Bridging Language and Items for Retrieval and Recommendation: Benchmarking LLMs as Semantic Encoders](https://mlsanity.com/paper/2403.03952/) (2024)
- [AnnoLLM: Making Large Language Models to Be Better Crowdsourced Annotators](https://mlsanity.com/paper/2303.16854/) (2023)
- [Participatory Research for Low-resourced Machine Translation: A Case Study in African Languages](https://mlsanity.com/paper/2010.02353/) (2020)
- [Personalized Machine Translation: Preserving Original Author Traits](https://mlsanity.com/paper/1610.05461/) (2016)
- [Break the Checkbox: Challenging Closed-Style Evaluations of Cultural Alignment in LLMs](https://mlsanity.com/paper/2502.08045/) (2025)
- [Translating Across Cultures: LLMs for Intralingual Cultural Adaptation](https://mlsanity.com/paper/2406.14504/) (2024)
- [BanglaBook: A Large-scale Bangla Dataset for Sentiment Analysis from Book Reviews](https://mlsanity.com/paper/2305.06595/) (2023)
- [DEPTWEET: A Typology for Social Media Texts to Detect Depression Severities](https://mlsanity.com/paper/2210.05372/) (2022)
- [Cross-Cultural Differences in Mental Health Expressions on Social Media](https://mlsanity.com/paper/2402.11477/) (2024)
- [Measuring Sentiment Bias in Machine Translation](https://mlsanity.com/paper/2306.07152/) (2023)
- [Towards Style Alignment in Cross-Cultural Translation](https://mlsanity.com/paper/2507.00216/) (2025)

## Built on by (most cited first, 2 of 2)

- [XCR-Bench: Benchmarking Cross-Cultural Reasoning in LLMs via Culture-Specific Items and Hall's Triad](https://mlsanity.com/paper/2601.14063/) (2026)
- [BETA-Labeling for Multilingual Dataset Construction in Low-Resource IR](https://mlsanity.com/paper/2602.14488/) (2026)

## Abstract (arXiv)

Machine Translation (MT) is widely employed to address resource scarcity in low-resource languages by generating synthetic data from high-resource counterparts. While sentiment preservation in translation has long been studied, a critical but underexplored factor is the role of cultural alignment between source and target languages. In this paper, we hypothesize that semantic labels are drifted or altered during MT due to cultural divergence. Through a series of experiments across culturally sensitive and neutral domains, we establish three key findings: (1) MT systems, including modern Large Language Models (LLMs), induce label drift during translation, particularly in culturally sensitive domains; (2) unlike earlier statistical MT tools, LLMs encode cultural knowledge, and leveraging this knowledge can amplify label drift; and (3) cultural similarity or dissimilarity between source and target languages is a crucial determinant of label preservation. Our findings highlight that neglecting cultural factors in MT not only undermines label fidelity but also risks misinterpretation and cultural conflict in downstream applications.