Papermesa-optimization论文
Risks from Learned Optimization
A 2019 AI safety paper by Evan Hubinger et al. that systematically articulates the risks of deceptive alignment and mesa-optimization. It is an important reference in AI alignment research.
Core Facts
Timeline (last 90 days)
Sep 2
The concept of Deceptive Alignment was systematically articulated by Evan Hubinger and other researchers in the 2019 paper "Risks from Learned Optimization."
Verified65%