Summary of A Neuro-symbolic Approach to Multi-agent Rl For Interpretability and Probabilistic Decision Making, by Chitra Subramanian and Miao Liu and Naweed Khan and Jonathan Lenchner and Aporva Amarnath and Sarathkrishna Swaminathan and Ryan Riegel and Alexander Gray
A Neuro-Symbolic Approach to Multi-Agent RL for Interpretability and Probabilistic Decision Making
by Chitra Subramanian, Miao Liu, Naweed Khan, Jonathan Lenchner, Aporva Amarnath, Sarathkrishna Swaminathan, Ryan Riegel, Alexander Gray
First submitted to arxiv on: 21 Feb 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Neural and Evolutionary Computing (cs.NE)
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 introduces an event-driven formulation of multi-agent reinforcement learning (MARL) that addresses issues of interpretability, sample efficiency, and partial observability in real-world applications. The authors propose a neuro-symbolic approach using Logical Neural Networks (LNN) as a function approximator to train rules-based policies that are both logical and interpretable. To enable decision-making under uncertainty, the authors develop Probabilistic Logical Neural Networks (PLNN), which combines logical reasoning with probabilistic graphical models. PLNN nodes form the unifying element combining probabilistic logic and Bayes Nets, permitting inference for variables with unobserved states. The paper demonstrates these contributions in a system-on-chip application. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper talks about a way to make machines learn together in complex systems. They’re trying to solve some big problems that happen when many agents (like robots or computers) work together and need to make decisions. They propose a new method called Probabilistic Logical Neural Networks (PLNN), which is like a special kind of neural network that can reason logically and make smart decisions even when things are uncertain. |
Keywords
» Artificial intelligence » Inference » Neural network » Reinforcement learning
