Metraipolis › archive › dc_eb8268115b79

Same Values, Different Languages? From Multilingual Probing to Steering LLMs Toward Chinese Social Values — arXiv:2609.08515

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.08515
License (as archived)
CC0-1.0
Archived
2026-10-16T12:48:58.727Z
SHA-256 of the archived text
55e016f7e319cc000e28060c1a40d22ad34d62314bf33fb1864ad565e3458777
Stable URL of this page
https://metraipolis.com/documento/dc_eb8268115b79

How to cite

Same Values, Different Languages? From Multilingual Probing to Steering LLMs Toward Chinese Social Values — arXiv:2609.08515 — arXiv (en), https://arxiv.org/abs/2609.08515 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eb8268115b79 (Metraipolis archive dc_eb8268115b79, sha256 55e016f7e319cc000e28060c1a40d22ad34d62314bf33fb1864ad565e3458777)
BibTeX
@misc{dceb8268115b79,
  title = {Same Values, Different Languages? From Multilingual Probing to Steering LLMs Toward Chinese Social Values — arXiv:2609.08515},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.08515},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eb8268115b79; sha256 55e016f7e319cc000e28060c1a40d22ad34d62314bf33fb1864ad565e3458777},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.08515. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As Large Language Models (LLMs) are increasingly integrated into human society, aligning them with pluralistic social values has become a critical priority. However, whether LLMs exhibit consistent value preferences across languages remains underexplored, particularly for culturally grounded values, which are more abstract and difficult to evaluate and align than safety-centric principles. We investigate this issue through Chinese Social Values (CSV), a value system rooted in Chinese culture and comprising $12$ dimensions across national, societal, and personal levels. We construct C-Voices, the first comprehensive multilingual contrastive probe dataset for CSV, with 86,400 dilemma-based instances in six languages, each pairing a CSV-aligned action with a value-conflicting alternative. Building on the contrastive probes of C-Voices, we then propose a fine-tuning-free value vector steering method that derives value directions from hidden-state discrepancies and selectively intervenes on value-sensitive layers during inference. Experiments on six languages show that CSV-oriented preferences are model-dependent and language-sensitive, with the same dilemma eliciting divergent responses across languages. Our method achieves effective CSV steering, supports cross-lingual transfer of value vectors, and generalizes to existing FLAMES and ValuePrism.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-09 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.08515 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T12:30:25.185Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).