1 paper
Jiho Lee, Bingquan Zhang, Lin Ting Liang +2
Hallucination stress tests can yield similar average scores across language models even when different prompts elicit hallucinations in each model. We develop a procedure that iter…