CommunityBench: Benchmarking Community-Level Alignment across Diverse Groups and Tasks
This addresses the challenge of scalable and pluralistic alignment for AI systems, though it is incremental by building on existing alignment strategies.
The authors tackled the problem of aligning large language models with human values by proposing community-level alignment as a middle ground between universal and individual approaches, and introduced CommunityBench, a benchmark that reveals current models have limited capacity to model community-specific preferences.
Large language models (LLMs) alignment ensures model behaviors reflect human value. Existing alignment strategies primarily follow two paths: one assumes a universal value set for a unified goal (i.e., one-size-fits-all), while the other treats every individual as unique to customize models (i.e., individual-level). However, assuming a monolithic value space marginalizes minority norms, while tailoring individual models is prohibitively expensive. Recognizing that human society is organized into social clusters with high intra-group value alignment, we propose community-level alignment as a "middle ground". Practically, we introduce CommunityBench, the first large-scale benchmark for community-level alignment evaluation, featuring four tasks grounded in Common Identity and Common Bond theory. With CommunityBench, we conduct a comprehensive evaluation of various foundation models on CommunityBench, revealing that current LLMs exhibit limited capacity to model community-specific preferences. Furthermore, we investigate the potential of community-level alignment in facilitating individual modeling, providing a promising direction for scalable and pluralistic alignment.