1 paper · 1 filter
Songrun He, Linying Lv, Asaf Manela +1
Large language models are increasingly used in social sciences, but their training data can introduce lookahead bias and training leakage. A good chronologically consistent languag…