Summary of Rewards Encoding Environment Dynamics Improves Preference-based Reinforcement Learning, by Katherine Metcalf and Miguel Sarabia and Barry-john Theobald
Rewards Encoding Environment Dynamics Improves Preference-based Reinforcement Learning
by Katherine Metcalf, Miguel Sarabia, Barry-John Theobald
First submitted to arxiv on: 12 Nov 2022
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary In this paper, researchers introduce a new method called REED (Reward Encoding for Environment Dynamics) to reduce the number of labels required in preference-based reinforcement learning algorithms. By encoding environment dynamics in the reward function, REED enables state-of-the-art preference-based RL frameworks to learn from fewer preference labels. The authors hypothesize that REED better partitions the state-action space and facilitates generalization to unlabelled state-action pairs. They explore the benefits of REED within two popular preference learning frameworks (PrefPPO and PEBBLE) and demonstrate improved policy learning speed and performance across various experimental conditions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In this paper, scientists find a way to make machines learn faster and better by using human feedback. They created a new method called REED that helps machines learn from fewer examples of what’s good or bad. This makes it easier for humans to teach machines without needing too many labels. The researchers tested their idea in two different ways and showed that it works better than usual methods. |
Keywords
* Artificial intelligence * Generalization * Reinforcement learning
