§ feed · storyline
Learning to model other minds
DeepMind releases LOLA, a multi-agent learning algorithm that accounts for other agents' learning dynamics and discovers collaborative strategies such as tit-for-tat in iterated games.
We’re releasing an algorithm which accounts for the fact that other agents are learning too, and discovers self-interested yet collaborative strategies like tit-for-tat in the iterated prisoner’s dilemma.
This algorithm, Learning with Opponent-Learning Awareness (LOLA), is a small step towards agents that model other minds.
§ sources1 publication · timeline below
- openai.comLearning to model other mindsprimary