MMODELYST
People/Yongxiang Li
PPL

Yongxiang Li

Papers (2)
Prompt-Level Reward Specifications for Open-Ended Post-TrainingPlant, Persist, Trigger: Sleeper Attack on Large Language Model Agents