Scalable agent alignment via reward modeling: a research direction.
Jan LeikeDavid KruegerTom EverittMiljan MarticVishal MainiShane LeggPublished in: CoRR (2018)
Keyphrases
- multi agent
- multi agent systems
- learning agent
- decision making
- reinforcement learning
- multiagent systems
- social simulation
- data sets
- image alignment
- reward function
- intelligent agents
- autonomous agents
- software agents
- knowledge base
- mobile agents
- modeling language
- action selection
- lightweight
- dynamic programming
- agent systems
- multiple sequence alignment
- neural network