News for hackers. Stuff that matters.
Re:OpenAI Five by YeGoblynQueenne on Monday June 25, 2018 @01:16PM (#17393998)
>> Our system learns to generalize behaviors in a sophisticated way.

Could you elaborate? One of the criticisms of RL and statistical machine learning in general is that models generalise extremely poorly, unless provided with unrealistic amounts of training data.

There are no comments on this comment yet.