1 paper · 1 filter
Trung Nguyen Quang, Yiming Gao, Fanyi Pu +3
When an omnimodal large language model accepts a question whose textual premise contradicts what it actually sees or hears, does the failure lie in perception or in action? Recent…