Gave an invited talk in Tokyo on Investigating differences in content safety of Large Language Models.