Pith. sign in

REVIEW 2 cited by

Speaking Multiple Languages Affects the Moral Bias of Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2211.07733 v2 pith:ERNYAFLJ submitted 2022-11-14 cs.CL

Speaking Multiple Languages Affects the Moral Bias of Language Models

classification cs.CL
keywords modelslanguagesmoralenglishlanguagepmlmsapplycomparing
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Pre-trained multilingual language models (PMLMs) are commonly used when dealing with data from multiple languages and cross-lingual transfer. However, PMLMs are trained on varying amounts of data for each language. In practice this means their performance is often much better on English than many other languages. We explore to what extent this also applies to moral norms. Do the models capture moral norms from English and impose them on other languages? Do the models exhibit random and thus potentially harmful beliefs in certain languages? Both these issues could negatively impact cross-lingual transfer and potentially lead to harmful outcomes. In this paper, we (1) apply the MoralDirection framework to multilingual models, comparing results in German, Czech, Arabic, Chinese, and English, (2) analyse model behaviour on filtered parallel subtitles corpora, and (3) apply the models to a Moral Foundations Questionnaire, comparing with human responses from different countries. Our experiments demonstrate that, indeed, PMLMs encode differing moral biases, but these do not necessarily correspond to cultural differences or commonalities in human opinions. We release our code and models.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. CURE-Med: Curriculum-Informed Reinforcement Learning for Multilingual Medical Reasoning

    cs.AI 2026-01 unverdicted novelty 6.0

    CURE-MED pairs a new 13-language medical reasoning benchmark with curriculum RL to raise logical correctness to 70% and language consistency to 95% at 32B scale while outperforming baselines.

  2. Camellia: Benchmarking Cultural Biases in LLMs for Asian Languages

    cs.CL 2025-10 conditional novelty 6.0

    Across nine Asian languages, multilingual LLMs favor Western cultural entities in 30-40% of culturally grounded contexts, with model-specific sentiment biases and extraction accuracy gaps.