1 paper · 1 filter
Obed Junias, Maria Leonor Pacheco
Large language models often fail when answer options require combining atomic judgments under explicit logical operators, even when they judge the individual atoms correctly. We st…