Lu Lu

A staggering 31.85% of model responses in pulmonary nodule assessments exhibited imaging hallucinations, revealing alarming safety risks in AI-driven clinical decision-making.

GPT-5.4 outshines other LLMs in cardiovascular care tasks, but major performance gaps reveal urgent areas for improvement in clinical AI applications.

NOTES reduces design dimensionality from 256 to 25 while achieving over 95% efficiency in inverse design tasks, outpacing conventional methods.

FlexiSLM can operate at frame rates as low as 4.0 Hz while maintaining high-quality speech, effectively halving inference time compared to traditional models.

Strong task performance in clinical AI models can mask significant process instabilities, particularly in contradiction detection and hallucination management.