The third AI Safety Camp took place in April 2019 in Madrid. Our teams have worked on exciting projects which are summarized below:
Categorizing Wireheading in Partially Embedded Agents:
AI Safety Debate and Its Applications:
Team: Debate – Vojta Kovarik, Anna Gajdova, David Lindner, Lukas Finnveden, Rajashree Agrawal
Regularization and visualization of attention in reinforcement learning agents
Team: Dmitry Nikulin, Sebastian Kosch, Fabian Steuer, Hoagy Cunningham
Read our research report here.