Weida Wang

Adversarial purification can be dramatically improved by focusing on patch-level semantics, leading to state-of-the-art performance in defending against adversarial attacks.

A single structural edit can drastically impair LLM performance in molecular tasks, highlighting the fragility of their generalization capabilities.

PhySciBench reveals that top LLMs struggle with scientific reasoning, achieving only 33.5% accuracy, while DelveAgent demonstrates a promising 7.5% improvement in performance.