Seong Joon Oh

large language models · LLM evaluation · LLM agents · reasoning · vision-language models · compositional understanding · CLIP · out-of-distribution generalisation · shortcut learning · spurious correlations · trustworthy machine learning · AI safety · robustness · uncertainty quantification · explainability · interpretability · feature attribution · privacy in machine learning · training data attribution · membership inference · deep learning · computer vision · neural networks