Document-level relation extraction via self-correcting prompt optimization and preference alignment
This paper proposes SPO-PA, which combines self-correcting prompt optimization (SPO) and preference alignment (PA), which compares LLM predictions with reference annotations and applies source-specific constraints to these signals.
Fu-Yuan Wei, Wen-Zhong Yang, Tiquan Gu et al.
· Journal of King Saud Univers... · 0 citations