Week 3 Updates!
Hello everyone, and welcome to Week 3 of my Senior Research Project on Reinforcement Learning!
This week was a bit different due to the senior trip, which meant I didn’t have as much time for hands-on coding. Instead, I focused on deepening my theoretical understanding of reinforcement learning by reading through a fascinating research paper and experimenting with an open-source reinforcement learning model to get a better visual grasp of how these systems work in action.
One of the key papers I went through was “Interpretable and Explainable Logical Policies via Neurally Guided Symbolic Abstraction” (NeurIPS 2023). This research explores how symbolic policies—which are more transparent and human-readable—can be combined with neural networks to make reinforcement learning models both powerful and explainable. The idea is that while deep learning models are great at decision-making, they often act as “black boxes,” making it difficult to understand why they make certain choices. By introducing symbolic reasoning, the goal is to create policies that can be logically interpreted and even manually tweaked, which could be useful in critical applications like robotics, healthcare, and finance.
Alongside reading, I also tried running an open-source reinforcement learning model to see how these systems behave in real time. The goal wasn’t to tweak anything yet but just to visualize the training process, see how rewards influence learning, and observe how different models converge to optimal strategies. Reinforcement learning can sometimes feel very abstract when learning from equations alone, so actually seeing an agent learn from its environment made the concepts much clearer.
Next week, I plan to dive deeper into coding my own small reinforcement learning experiments and start working with Bellman’s Equation to further develop my understanding of how agents optimize their decision-making. Stay tuned for more updates!
Would love to hear any feedback or suggestions—until next time!
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.