2 papers
cs.CL2026
On the Robustness of Knowledge Editing for Detoxification
Ming Dong, Shiyi Tang, Ziyan Peng +2
Knowledge-Editing-based (KE-based) detoxification has emerged as a promising approach for mitigating harmful behaviours in Large Language Models. Existing evaluations, however, lar…
cs.LG2024
STEMO: Early Spatio-temporal Forecasting with Multi-Objective Reinforcement Learning
Wei Shao, Yufan Kang, Ziyan Peng +4
Accuracy and timeliness are indeed often conflicting goals in prediction tasks. Premature predictions may yield a higher rate of false alarms, whereas delaying predictions to gathe…