Summary of Learning Successor Features the Simple Way, by Raymond Chua et al.
Learning Successor Features the Simple Way
by Raymond Chua, Arna Ghosh, Christos Kaplanis, Blake A. Richards, Doina Precup
First submitted to arxiv on: 29 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 The proposed method learns representations that do not exhibit catastrophic forgetting or interference in non-stationary environments using Successor Features (SFs) in Deep Reinforcement Learning (RL). The authors introduce a novel approach that combines a Temporal-difference (TD) loss and a reward prediction loss to capture the basic mathematical definition of SFs. This method is efficient, matching or outperforming existing SF learning techniques in various scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way has been found to learn representations that don’t forget or get mixed up when things change in a situation. This helps with Deep Reinforcement Learning (RL) and Successor Features (SFs). The idea is simple: use two types of losses, one that learns from what happened before and one that predicts rewards. This method works well for finding paths in mazes and reaching high levels quickly. |
Keywords
* Artificial intelligence * Reinforcement learning
