Exploring Multilingual Concepts of Human Value in Large Language Models: Is Value Alignment Consistent, Transferable and Controllable across Languages?

Domain:

natural language processing

Record type:

paper

Creator:

Xu,DonGuoWu,

Host:

Prior research has revealed that certain abstract concepts are linearly represented as directions in the representation space of LLMs, predominantly centered around English. In this paper, we extend this investigation to a multilingual context, with a specific focus on human values-related concepts (i.e., value concepts) due to their significance for AI safety. Through our comprehensive exploration covering 7 types of human values, 16 languages and 3 LLM series with distinct multilinguality (e.g., monolingual, bilingual and multilingual), we first empirically confirm the presence of value concepts within LLMs in a multilingual format. Further analysis on the cross-lingual characteristics of these concepts reveals 3 traits arising from language resource disparities: cross-lingual inconsistency, distorted linguistic relationships, and unidirectional cross-lingual transfer between high- and low-resource languages, all in terms of value concepts. Moreover, we validate the feasibility of cross-lingual control over value alignment capabilities of LLMs, leveraging the dominant language as a source language. Ultimately, recognizing the significant impact of LLMs' multilinguality on our results, we consolidate our findings and provide prudent suggestions on the composition of multilingual data for LLMs pre-training. EMNLP 2024 findings, code&dataset: github.com

Visit

arxiv.org

Evaluating Cross-National Value Alignment in Large Language Models: Challenges and Limitations of Survey-Based Approaches

Large language models (LLMs) often exhibit cultural biases, raising questions about their alignment

MEGAVERSE: Benchmarking Large Language Models Across Languages, Modalities, Models and Tasks

There has been a surge in LLM evaluation research to understand LLM capabilities and limitations. Ho

Sycophancy as a Multilingual Alignment Failure: How Safety Degrades Across Languages, Topics, and Models

Safety-aligned large language models often exhibit sycophancy, which is the tendency to affirm users

Investigating Cultural Alignment of Large Language Models

The intricate relationship between language and culture has long been a subject of exploration withi

Multilingual Prompt Engineering in Large Language Models: A Survey Across NLP Tasks

Large language models (LLMs) have demonstrated impressive performance across a wide range of Natural

Are Knowledge and Reference in Multilingual Language Models Cross-Lingually Consistent?

Cross-lingual consistency should be considered to assess cross-lingual transferability, maintain the