Search papers, labs, and topics across Lattice.
This paper introduces "sensitive prompts" as a novel method for evaluating fairness in LLMs, focusing on inputs that, while not inherently biased, are likely to elicit biased or inadequate responses due to their sensitive content. They construct SensY, a dataset of 12,801 prompts categorized as sensitive and non-sensitive across seven domains, and evaluate the responses of three open-source LLMs. The results demonstrate that LLMs often fail to address the ethical and contextual implications of sensitive prompts, highlighting the utility of prompt sensitivity as an early-warning mechanism for fairness risks.
LLMs can ace the facts but still whiff on the ethics, relational context, and broader implications when answering seemingly innocuous "sensitive" prompts.
Large Language Models (LLMs) are being increasingly integrated into software systems, offering powerful capabilities but also raising concerns about fairness. Existing fairness benchmarks, however, focus on stereotype-specific associations, which limit their ability to anticipate risks in diverse, real-world contexts. In this paper, we propose sensitive prompts as a new abstraction for fairness evaluation: inputs that are not inherently biased but are more likely to elicit biased or inadequate responses due to the sensitivity of their content. We construct and release SensY, a dataset of 12,801 prompts, categorized as sensitive and non-sensitive, spanning seven thematic domains, combining synthetic generation and real user inputs. Using this dataset, we query three open-source LLMs and manually analyze 4,500 responses to evaluate their adequacy in answering sensitive prompts. Results show that while models often provide factually correct answers, they frequently fail to acknowledge the ethical, relational, or contextual implications of sensitive inputs. In addition, we develop an automated classifier for predicting prompt sensitivity, achieving robust performance on sensitive prompts. Our findings demonstrate that prompt sensitivity can serve as an effective early-warning mechanism for fairness risks in LLMs. This perspective shifts fairness assessment from reactive mitigation toward preventive design, enabling developers to anticipate and manage bias before deployment.
