1 paper
Keren Fuentes, Aaron Mueller
Language models (LMs) often pass behavioral bias evaluations, but it remains unclear whether they no longer represent the underlying associations that give rise to biases, or have…