Hacker Next
Large language models develop novel social biases through adaptive explorationopenreview.net
181 points by paimapi 15 hours ago | 95 comments