一条鱼，两条鱼，但并非整个海洋：对齐降低了语言模型的概念多样性
One fish, two fish, but not the whole sea: Alignment reduces language models' conceptual diversity
AI总结：
本研究提出一种衡量LLM群体概念多样性的新方法，发现对齐模型比指令微调模型多样性更低，揭示了价值对齐与概念多样性之间的权衡。
摘要：
社会科学和心理学领域的研究者最近提出使用大型语言模型（LLMs）作为行为研究中人类的替代品。除了关于LLMs是否准确捕捉人口层面模式的争论之外，这也引发了关于LLMs是否能捕捉类似人类的概念多样性的问题。另外，关于训练后的对齐（RLHF或RLAIF）是否影响模型内部多样性也存在争议。受人类研究的启发，我们采用一种新方法来衡量合成生成的LLM“群体”的概念多样性，该方法将模拟个体的内部变异性与群体层面的变异性联系起来。我们使用这种方法在两个具有丰富人类行为数据的领域评估未对齐和对齐的LLMs。虽然没有任何模型达到类似人类的多样性，但对齐模型通常表现出比其指令微调对应模型更少的多样性。我们的研究结果突显了增加模型价值对齐与减少其概念表征多样性之间潜在的权衡。
英文摘要：
Researchers in social science and psychology have recently proposed using large language models (LLMs) as replacements for humans in behavioral research. In addition to arguments about whether LLMs accurately capture population-level patterns, this has raised questions about whether LLMs capture human-like conceptual diversity. Separately, it is debated whether post-training alignment (RLHF or RLAIF) affects models' internal diversity. Inspired by human studies, we use a new way of measuring the conceptual diversity of synthetically-generated LLM "populations" by relating the internal variability of simulated individuals to the population-level variability. We use this approach to evaluate non-aligned and aligned LLMs on two domains with rich human behavioral data. While no model reaches human-like diversity, aligned models generally display less diversity than their instruction fine-tuned counterparts. Our findings highlight potential trade-offs between increasing models' value alignment and decreasing the diversity of their conceptual representations.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
