Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment Paper • 2401.12474 • Published Jan 23, 2024 • 36
