Cost-aware Discovery of Contextual Failures using Bayesian Active Learning
Conference on Robot Learning (CoRL), 2025
Existing methods for failure discovery rely on explicitly defined analytical cost functions to characterize failures, often overlooking the underlying causes and diversity of discovered failure scenarios. In this work, we propose a novel failure discovery framework that integrates contextual reasoning from an expert (human/LLM) in the failure analysis process, specifically tailored for high evaluation-cost applications.
Our method incorporates expert-in-the-loop feedback to construct a probabilistic surrogate model of failures using Bayesian inference. This model is iteratively refined and leveraged to guide an active learning strategy that prioritizes the discovery of diverse failure cases. We empirically validate our approach across a range of tasks for high-cost contextual falsification in robotic manipulation and autonomous driving. The project website can be found here.
Recommended citation: A. Parashar, J. Zhang, Y. Li, C. Fan (2025). "Cost-aware Discovery of Contextual Failures using Bayesian Active Learning." 9th Conference on Robot Learning.
Download Paper