§ feed · storyline
Meta-learning for wrestling
Meta-learning agent defeats stronger non-meta-learning opponents in simulated robot wrestling and adapts to physical malfunction, according to new research.
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.
§ sources1 publication · timeline below
- openai.comMeta-learning for wrestlingprimary