Meta-learning for wrestling
Meta-learning agent that rapidly adapts to defeat stronger opponents and physical malfunctions in simulated robot wrestling.
Excerpt
We show that for the task of simulated robot wrestling, a meta-learning agent can learn to quickly defeat a stronger non-meta-learning agent, and also show that the meta-learning agent can adapt to physical malfunction.
Read at source: https://openai.com/index/meta-learning-for-wrestling