Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment.

Keming Lu, Bowen Yu, Chang Zhou, Jingren Zhou

Browse the full ACL paper archive.