Recent Activities
Reinforcement Learning with an old acquaintance
I’m revisiting an old passion: reinforcement learning—this time by training an agent to play Super Mario. The project is a hands-on playground for RL fundamentals (reward design, exploration, and learning stability). Code snippets and notes are available on GitHub.
And while the agent is playing by itself, you can go for a hike with alpacas or look up for northern lights...
Wanna see an aurora?
After missing a few opportunities, I built a small aurora forecast that combines multiple data sources to highlight promising nights. It’s designed to support planning—so you spend less time guessing and hopefully more time outside looking up. Try the app.
Tiny Worlds in Hamburg
After nearly 20 years, I returned to Miniatur Wunderland in Hamburg—still astonishing in its level of detail. This is a fine place where craftsmanship and storytelling meet in miniature. I can recommend a late visit with fewer people around you. If you like a challenge, try to find 8 or more alien (groups) - and please send your pictures.
Ever had an alpaca on the leash?
Hiking with an alpaca is a unique and delightful experience! We took a tour near Cologne in the "Bergisches Land." Alpacas are gentle (if the chemistry is right), curious, and fluffy animals, making them great companions for a hike. Just don’t expect them to carry your bags.
One in a Million
What does "one in a million" looks like? I figure it as a single red dot hidden among 999.999 gray ones. Create your own grids and get some assistance—or even the solution. Try the app.
I've also written Medium article on that topic: One In A Million — Once In A Lifetime: Developing a sense of probabilities with visualizations
March 14th
Before May the 4th, winter has to end—so we celebrate Pi Day (March 14th) with pie, not cake. This one was baked by my older daughter and cut into perfectly even twelfths.😋
