Cross-Lingual Stability of LLM Judges Under Controlled Generation: Evidence from Finno-Ugric Languages

Isaac Chung, Linda Freienthal


Abstract
Cross-lingual evaluation of large language models (LLMs) typically conflates two sources of variance: genuine model performance differences and measurement instability. We investigate evaluation reliability by holding generation conditions constant while varying target language. Using synthetic customer-support dialogues generated with identical parameters across Estonian, Finnish, and Hungarian, we test whether automatic metrics and LLM-as-a-judge scoring produce stable model rankings across these morphologically rich, related Finno-Ugric languages. With a small set of Estonian native speaker annotations as a reference point, we find systematic ranking instabilities: surface-level metrics (lexical diversity, surface and semantic similarity) maintain cross-language stability, but pragmatic judgments (coherence, instruction-following) exhibit rank inversions and near-zero correlations. Because generation is controlled, these inconsistencies reflect how judge scoring behaves differently across languages rather than true model differences.This controlled design provides a diagnostic probe: evaluation methods that fail to maintain stability under identical generation conditions signal transfer failure before deployment. Our findings suggest that zero-shot judge transfer is unreliable for discourse-level assessment in morphologically rich languages, motivating language-specific calibration against targeted human baselines. We release our controlled generation protocol, synthetic data, and evaluation framework to enable replication across language families at https://github.com/isaac-chung/cross-lingual-stability-judges.
Anthology ID:
2026.mme-main.8
Volume:
Proceedings of the First Workshop on Multilingual Multicultural Evaluation
Month:
March
Year:
2026
Address:
Rabat, Morocco
Editors:
Pinzhen Chen, Vilém Zouhar, Hanxu Hu, Simran Khanuja, Wenhao Zhu, Barry Haddow, Alexandra Birch, Alham Fikri Aji, Rico Sennrich, Sara Hooker
Venues:
MME | WS
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
133–148
Language:
URL:
https://aclanthology.org/2026.mme-main.8/
DOI:
Bibkey:
Cite (ACL):
Isaac Chung and Linda Freienthal. 2026. Cross-Lingual Stability of LLM Judges Under Controlled Generation: Evidence from Finno-Ugric Languages. In Proceedings of the First Workshop on Multilingual Multicultural Evaluation, pages 133–148, Rabat, Morocco. Association for Computational Linguistics.
Cite (Informal):
Cross-Lingual Stability of LLM Judges Under Controlled Generation: Evidence from Finno-Ugric Languages (Chung & Freienthal, MME 2026)
Copy Citation:
PDF:
https://aclanthology.org/2026.mme-main.8.pdf