AI Alignment with Changing and Influenceable Reward Functions.
Micah CarrollDavis FooteAnand SiththaranjanStuart RussellAnca D. DraganPublished in: CoRR (2024)
Keyphrases
- reward function
- artificial intelligence
- markov decision processes
- state space
- inverse reinforcement learning
- optimal policy
- reinforcement learning algorithms
- reinforcement learning
- expert systems
- state variables
- multiple agents
- machine learning
- markov decision process
- simple examples
- transition probabilities
- policy search
- generative model
- initially unknown
- dynamic programming
- data mining