作者位次 #4/6ACL 2026Poster· Natural Language Processing
CHOIR : Harmonizing Structured Persona Diversity for Robust Collaborative LLM Reasoning
Xiangjue Dong · Cong Wang · Maria Teleki · Millennium Bismay · Ruihong Huang · James Caverlee
摘要
Persona-assigned Large Language Models can adopt diverse roles, enabling personalized and context-aware reasoning. However, even minor demographic perturbations in personas, such as simple pronoun swaps, can alter reasoning trajectories, leading to divergent sets of correct answers on reasoning b...