CoSToM: Causal-oriented Steering for Intrinsic Theory-of-Mind Alignment in Large Language Models.

Mengfan Li, Xuanhua Shi, Yang Deng

Browse the full ACL paper archive.