AI Insight
This study addresses the Rashomon effect in machine learning, where multiple equally accurate models make different predictions for the same inputs. The researchers developed a method to audit ensemble decision systems by combining ensemble margin with local prediction variability to identify cases requiring human review. Through experiments on transformer models for natural language tasks and large language models for tabular data, they demonstrated that ensembles from the Rashomon set substantially reduce the risk of incorrect predictions going undetected compared to single models, while requiring only moderately more human interventions.
Why it matters
This work provides a practical framework for improving the reliability of AI decision systems in high-stakes applications like healthcare, finance, or criminal justice, where incorrect automated decisions have serious consequences. The ability to efficiently identify uncertain predictions using modest-sized ensembles makes the approach feasible for real-world deployment while significantly reducing risk.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Abstract: The Rashomon effect is a machine learning phenomenon where equally accurate models produce different predictions for the same inputs (predictive multiplicity). Existing work primarily focuses on multiplicity within individual models, but in more complex decision systems, the impact of the Rashomon effect is less well understood. In this work, we study multiplicity from the perspective of auditing incorrect ensemble predictions, where the decision to divert an instance for human review is based on a consistency criterion that combines the ensemble margin with a measure of local prediction variability for each constituent model. With mild assumptions about stability and smoothness, we show that the consistency scores of finite ensembles converge to the corresponding consistency score of the expected model from the Rashomon set as the ensemble size and the number of samples used to measure local prediction variability increase. To demonstrate the efficacy of the proposed criterion, we evaluate the framework with respect to transformer models applied to natural language understanding tasks and parameter-efficient fine-tuning of large language models used for tabular data classification tasks. Our experiments show that ensembling models from the Rashomon set substantially reduces the risk of incorrect predictions going unchecked compared with auditing a single model, while incurring only a moderate increase in the number of diversions. Moreover, the auditing behavior of the full Rashomon set can be closely approximated by finite ensembles of relatively modest size, with the risk approaching zero for some datasets. We further demonstrate that the proposed measure exhibits stronger agreement with established predictive multiplicity metrics than existing consistency measures, providing a more reliable way to capture multiplicity in the Rashomon set.