WVS Webinar on May 26. Can Language Models Reason about Individualistic Human Values and Preferences?
Registration: https://us02web.zoom.us/meeting/register/YmX3iLaFRDWvsJSS0uBKlw
Can Language Models Reason about Individualistic Human Values and Preferences?
Recent calls for pluralistic alignment emphasize that AI systems should address the diverse needs of all people. Yet, efforts in this space often require sorting people into fixed buckets of pre-specified diversity-defining dimensions (e.g., demographics), risking smoothing out individualistic variations or even stereotyping. To achieve an authentic representation of diversity that respects individuality, we propose individualistic alignment. While individualistic alignment can take various forms, we introduce IndieValueCatalog, a dataset transformed from the influential World Values Survey (WVS), to study language models (LMs) on the specific challenge of individualistic value reasoning. Given a sample of an individual's value-expressing statements, models are tasked with predicting this person's value judgments in novel cases. With IndieValueCatalog, we reveal critical limitations in frontier LMs, which achieve only 55 % to 65% accuracy in predicting individualistic values. Moreover, our results highlight that a precise description of individualistic values cannot be approximated only with demographic information. We also identify a partiality of LMs in reasoning about global individualistic values, as measured by our proposed Value Inequity Index. Finally, we train a series of IndieValueReasoners to reveal new patterns and dynamics into global human values.
Liwei Jiang is a Research Scientist at NVIDIA and recently completed her Ph.D. in Computer Science & Engineering at the University of Washington, advised by Yejin Choi. Her research focuses on AI safety and alignment, natural language processing, and human-centered AI, with an emphasis on building AI systems that genuinely benefit humanity.
‹ back