Physics 5700 Advanced Physics Laboratory PRL-style reports from three long-term experiments throughout a semester. This includes: 2-Slit Interference in the Quantum Limit, Boltzmann Constant Measurement via Johnson Noise, and Visible Spectroscopy experiments.
Physics 5700 (Advanced Physics Laboratory) serves as the capstone experimental experience for senior Physics and Engineering students at Ohio State. Unlike earlier labs that focus on verifying known laws with “cookie-cutter” setups, 5700 pushes us into the realm of professional scientific inquiry. The course is built around multi-week “Long Experiments” that require not just technical execution, but rigorous data analysis, uncertainty propagation, and professional communication.
The ultimate goal is to move beyond “getting the right answer” and instead focus on quantifying exactly how well we know that answer. This meant mastering the Physical Review Letters (PRL) style of reporting—learning to be concise, active in voice, and visually precise with data representation. It’s a course that demands high-level teamwork in the lab and total independence in the analytical “write-up” phase.
As a part of a two-semester undergraduate Order of Magnitude (OoM) astronomy course, the second semester culminated in a group project where we created our own OoM problems. Our group created a problem titled “A Positive Disaster”, a problem about ripping electrons off the moon! An LLM was used to create this cover image.
Imagine an alien race has come to our solar system with an “electron capture” magnet. They position the magnet at the moon and start ripping away electrons from the moon, making it positively charged, and they conveniently don’t repel each other or attract other positive charges when doing so. As the moon continues to get more positively charged, it will eventually “explode” (become unbound).
(a) Calculate the number of electrons you have to remove before the moon explodes/becomes unbound. (b) Is this a sensible amount of electrons?
In this post I talk about the problem itself and the writing process.
Loading data for AI models can be RAM and time-intensive. For a larger project (HackAI 2025) with a big dataset, I ran an optimization test on a manual loader and a Tensorflow pipeline to see the variations in time and RAM. As expected, Tensorflow’s model is much more efficient.
While working on our HackAI 2025 project, we found that our original way of loading the image data for our multi-headed regression was incredibly slow, and was crashing some of our computers due to RAM usage. Eventually, we converted our pipeline over to one that utilizes the more-optimized tf.data.Dataset pipeline. For our own entertainment, we made plots comparing the old method with the new method.
