Psychometric Alignment: Capturing Human Knowledge Distributions via Language Models
This addresses the need for better simulation of human-like responses in decision-making applications like education and policy, though it is incremental in proposing a new metric rather than a paradigm shift.
The paper tackles the problem of quantifying how closely language models (LMs) mimic human knowledge distributions, rather than just accuracy, by introducing a 'psychometric alignment' metric based on Item Response Theory. It finds significant misalignment between LMs and humans across three domains, with smaller LMs and persona-based prompts improving alignment, and training on human data enhancing it variably.
Language models (LMs) are increasingly used to simulate human-like responses in scenarios where accurately mimicking a population's behavior can guide decision-making, such as in developing educational materials and designing public policies. The objective of these simulations is for LMs to capture the variations in human responses, rather than merely providing the expected correct answers. Prior work has shown that LMs often generate unrealistically accurate responses, but there are no established metrics to quantify how closely the knowledge distribution of LMs aligns with that of humans. To address this, we introduce "psychometric alignment," a metric that measures the extent to which LMs reflect human knowledge distribution. Assessing this alignment involves collecting responses from both LMs and humans to the same set of test items and using Item Response Theory to analyze the differences in item functioning between the groups. We demonstrate that our metric can capture important variations in populations that traditional metrics, like differences in accuracy, fail to capture. We apply this metric to assess existing LMs for their alignment with human knowledge distributions across three real-world domains. We find significant misalignment between LMs and human populations, though using persona-based prompts can improve alignment. Interestingly, smaller LMs tend to achieve greater psychometric alignment than larger LMs. Further, training LMs on human response data from the target distribution enhances their psychometric alignment on unseen test items, but the effectiveness of such training varies across domains.