Summary of Latent-predictive Empowerment: Measuring Empowerment Without a Simulator, by Andrew Levy et al.
Latent-Predictive Empowerment: Measuring Empowerment without a Simulator
by Andrew Levy, Alessandro Allievi, George Konidaris
First submitted to arxiv on: 15 Oct 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Machine Learning (cs.LG); Robotics (cs.RO)
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 This paper addresses the challenge of empowering agents to learn large skillsets, but in a scalable manner. Recent methods have learned diverse skills by maximizing mutual information between skills and states, but these approaches require a model of transition dynamics, which can be difficult to learn in realistic settings with high-dimensional and stochastic observations. The authors propose Latent-Predictive Empowerment (LPE), an algorithm that computes empowerment in a more practical manner. LPE learns large skillsets by maximizing an objective that replaces mutual information between skills and states, requiring only a simpler latent-predictive model instead of a full simulator of the environment. Empirical results show that LPE outperforms other model-based approaches to empowerment. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper’s main goal is to help agents learn many skills. Current methods can do this, but they need information about how things change over time. This can be hard to get in real-life situations with lots of data and randomness. The authors came up with a new way called Latent-Predictive Empowerment (LPE) that makes it easier to empower agents. LPE works by looking at skills and states and trying to find patterns. It does this without needing a full map of the environment, which is helpful in real-life situations. |
