FaithEval: Can Your Language Model Stay Faithful to Context, Even If the Moon is Made of Marshmallows?
Published in ICLR, 2025
Recommended citation: Yifei Ming, Senthil Purushwalkam, Shrey Pandit, Zixuan Ke, Xuan-Phi Nguyen, Caiming Xiong, Shafiq Joty (2025). FaithEval: Can Your Language Model Stay Faithful to Context, Even If the Moon is Made of Marshmallows? ICLR 2025.
Abstract
This work evaluates whether language models remain faithful to supplied context, including when that context conflicts with prior knowledge.