Summary of Visualpredicator: Learning Abstract World Models with Neuro-symbolic Predicates For Robot Planning, by Yichao Liang et al.
VisualPredicator: Learning Abstract World Models with Neuro-Symbolic Predicates for Robot Planning
by Yichao Liang, Nishanth Kumar, Hao Tang, Adrian Weller, Joshua B. Tenenbaum, Tom Silver, João F. Henriques, Kevin Ellis
First submitted to arxiv on: 30 Oct 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computer Vision and Pattern Recognition (cs.CV); 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 The paper proposes a novel abstraction language, Neuro-Symbolic Predicates, which combines symbolic and neural knowledge representations to enable broad intelligence in agents. The authors present an online algorithm for inventing these predicates and learning abstract world models. The approach is compared to hierarchical reinforcement learning, vision-language model planning, and symbolic predicate invention methods on simulated robotic domains. Results show improved sample complexity, out-of-distribution generalization, and interpretability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper creates a new way for machines to understand tasks by combining two types of knowledge: symbolic and neural. The authors developed an algorithm that helps machines learn abstract representations of the world. They tested their approach on robotic simulations and found it did better than other methods in terms of learning quickly, handling unfamiliar situations, and being easy to understand. |
Keywords
» Artificial intelligence » Generalization » Language model » Reinforcement learning
