CuMA: Aligning LLMs with Sparse Cultural Values via Demographic-Aware Mixture of Adapters
This work addresses the challenge of cultural alignment for LLMs serving global audiences, offering a novel solution to preserve diversity, though it is incremental in improving alignment methods.
The paper tackles the problem of aligning large language models with diverse cultural values by addressing Mean Collapse, where dense models converge to a generic average and fail to represent distinct groups; the result is that CuMA, a demographic-aware mixture of adapters, achieves state-of-the-art performance on benchmarks like WorldValuesBench, significantly outperforming baselines and mitigating mean collapse.
As Large Language Models (LLMs) serve a global audience, alignment must transition from enforcing universal consensus to respecting cultural pluralism. We demonstrate that dense models, when forced to fit conflicting value distributions, suffer from \textbf{Mean Collapse}, converging to a generic average that fails to represent diverse groups. We attribute this to \textbf{Cultural Sparsity}, where gradient interference prevents dense parameters from spanning distinct cultural modes. To resolve this, we propose \textbf{\textsc{CuMA}} (\textbf{Cu}ltural \textbf{M}ixture of \textbf{A}dapters), a framework that frames alignment as a \textbf{conditional capacity separation} problem. By incorporating demographic-aware routing, \textsc{CuMA} internalizes a \textit{Latent Cultural Topology} to explicitly disentangle conflicting gradients into specialized expert subspaces. Extensive evaluations on WorldValuesBench, Community Alignment, and PRISM demonstrate that \textsc{CuMA} achieves state-of-the-art performance, significantly outperforming both dense baselines and semantic-only MoEs. Crucially, our analysis confirms that \textsc{CuMA} effectively mitigates mean collapse, preserving cultural diversity. Our code is available at https://github.com/Throll/CuMA.