The remarkable capabilities of Large Language Models (LLMs) make them increasingly compelling for adoption in real-world healthcare applications. However, the risks associated with using LLMs in medical applications have not been systematically characterized. We propose using five key principles for safe and trustworthy medical AI: Truthfulness, Resilience, Fairness, Robustness, and Privacy, along with ten specific aspects. Under this comprehensive framework, we introduce a novel MedGuard benchmark with 1,000 expert-verified questions. Our evaluation of 11 commonly used LLMs shows that the current language models, regardless of their safety alignment mechanisms, generally perform poorly on most of our benchmarks, particularly when compared to the high performance of human physicians. Despite recent reports indicate that advanced LLMs like ChatGPT can match or even exceed human performance in various medical tasks, this study underscores a significant safety gap, highlighting the crucial need for human oversight and the implementation of AI safety guardrails.

In scientific machine learning, the task of identifying partial differential equations accurately from sparse and noisy data poses a significant challenge. Current sparse regression methods may identify inaccurate equations on sparse and noisy datasets and are not suitable for varying coefficients. To address this issue, we propose a hybrid framework that combines two alternating direction optimization phases: discovery and embedding. The discovery phase employs current well-developed sparse regression techniques to preliminarily identify governing equations from observations. The embedding phase implements a recurrent convolutional neural network (RCNN), enabling efficient processes for time-space iterations involved in discretized forms of wave equation. The RCNN model further optimizes the imperfect sparse regression results to obtain more accurate functional terms and coefficients. Through alternating update of discovery-embedding phases, essential physical equations can be robustly identified from noisy and low-resolution measurements. To assess the performance of proposed framework, numerical experiments are conducted on various scenarios involving wave equation in elastic/viscoelastic and homogeneous/inhomogeneous media. The results demonstrate that the proposed method exhibits excellent robustness and accuracy, even when faced with high levels of noise and limited data availability in both spatial and temporal domains.

Ground-motion model (GMM) is the basis of many earthquake engineering studies. In this study, a novel physics-informed symbolic learner (PISL) method based on the Nest Generation Attenuation-West2 database is proposed to automatically discover mathematical equation operators as symbols. The sequential threshold ridge regression algorithm is utilized to distill a concise and interpretable explicit characterization of complex systems of ground motions. In addition to the basic variables retrieved from previous GMMs, the current PISL incorporates two a priori physical conditions, namely, distance and amplitude saturation. GMMs developed using the PISL, an empirical regression method (ERM), and an artificial neural network (ANN) are compared in terms of residuals and extrapolation based on obtained data of peak ground acceleration and velocity. The results show that the inter- and intra-event standard deviations of the three methods are similar. The functional form of the PISL is more concise than that of the ERM and ANN. The extrapolation capability of the PISL is more accurate than that of the ANN. The PISL-GMM used in this study provide a new paradigm of regression that considers both physical and data-driven machine learning and can be used to identify the implied physical relationships and prediction equations of ground motion variables in different regions.