Peiyi Wang

PCA-Bench tests large language models in complex decision-making scenarios.

A new dataset aims to enhance AI's grasp of scientific images and reasoning.

Exploring how preference learning improves language model alignment with human expectations.