On January 22, 2024, I gave a winter meeting talk about the SKACH projects at ZHAW.
On December 6 2023, I gave a talk about the projects that I work on at ZHAW.
On October 30 2023, I presented a poster on the Hammer & Nails conference in Ascona, displaying our ongoing work for the SKACH consortium at the Centre for Artificial Intelligence at ZHAW. We’re using image-to-image translation cGANs with simulated maps of galaxies to infer astrophysical properties, which would otherwise stay hidden in observations by radio telescopes like the SKA… dark matter for instance.
These are the slides for a short talk with the title “Deep learning domain translation between mock observations and hydrodynamical simulations” at the Swiss SKA Days 2023 conference in Zurich. It presents a diagnosis for the reason we had a hard time training our generative image-to-image translation models on projection images of IllustrisTNG galaxies. In order to make the presentation easier to follow (and for archival), I used the reveal.js framework to turn my presentation slides into a HTML/javascript website.
These are the slides for a short talk with the title “Image-to-image translation between SPH simulations and SKA mocks” at the SKACH spring meeting on June 02 2023 in Geneva. It is meant as a status update of our research for SKACH at the Centre for Artificial Intelligence at ZHAW which started late summer in 2022. In order to make the presentation easier to follow (and for archival), I used the reveal.js framework to turn my presentation slides into a HTML/javascript website.
On Januar 11 2023, I presented a scientific poster displaying our ongoing work for the SKACH consortium at the Centre for Artificial Intelligence at ZHAW. Since long texts are never ideal on any physical presentation medium, I decided to upload a more detailed version of the poster here, for the extraordinarily interested parties…
These are the slides for a short talk with the title “Deep learning the mapping between SKA mock observations and hydrodynamical simulations” at the SKACH winter meeting on January 12 2023. It is meant as an introduction of our research for SKACH at the Centre for Artificial Intelligence at ZHAW which started late summer in 2022. In order to make the presentation easier to follow (and for archival), I used the reveal.js framework to turn my presentation slides into a HTML/javascript website.
\( \newcommand\cvec[1]{\begin{pmatrix}#1\end{pmatrix}} \newcommand\bra[1]{\langle#1\rvert} \newcommand\ket[1]{\lvert#1\rangle} \newcommand\braket[2]{\langle#1\rvert\,#2\rangle} \newcommand\ketbra[2]{\lvert#1\rangle\langle#2\rvert} \newcommand{\frasq}{\frac{1}{\sqrt{2}}} \) In my last post I demonstrated how easily a mathematical proof can lead to completely absurd conclusions if the "base truth" is incorrectly set up. This led me down the rabbit hole and I spent much of my leisure thinking about mathematical proofs in relation to artificial intelligence (AI), and related concepts like learning, understanding, computability, and consciousness. Rather unexpectantly, you might hear these concepts discussed in the context of quantum physics or even gravity. It is far more likely to hear opinions on such matters from neuroscientists, or philosophers, but being a "learned" physicist I cannot help it and would like to come at these subjects from an other perspective.
by Philipp Denzel and Prasenjit Saha In one of Einstein's notebooks from 1912, there is a curious figure with an even more curious caption. The caption (in German) reads Alle Dreiecke sind gleichschenklig meaning All triangles are isosceles. We guess a few historians of science have stared at the remark, shaken their heads in disbelief, and moved on. On the other hand, if you happen to have Grant Sanderson's video How to lie using visual proofs fresh in your mind, the sketch will look suspiciously familiar. There's no doubt, Albert and Grant are talking about the same fake proof! Here we'll explain, with the help of some interactive graphics, a mathematical sleight-of-hand hiding in Einstein's sketch. Spoiler alert: it will give away a punch line from the video.
On June 16 2022, I presented a short tech talk on “Recent Developments in High-Resolution Image Synthesis”. Since the reserved time for this talk was rather short, it is meant for people who want to catch up on the latest news in research around generative deep learning, and already are familiar with Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), Autoregressive models, etc. In order to make the presentation easier to follow (and for archival), I used the reveal.js framework to turn my presentation slides into a HTML/javascript website.
Some people hate it, some enjoy it… setting up a new machine, and a new operating system. Generally speaking, the majority of the people who enjoy this process are probably Linux users to which I count myself. There is even a word for the process of repeatedly setting up new machines in the Linux community… “distro-hopping”. After a while hopping from distro to distro though, I realized that the importance of distros is somewhat exaggerated in the community. One of the many advantages of Linux (at least if you are familiar with it) is the configurability and freedom to do whatever you want with your machine. This means you can make any distro look and feel like any other. In this post, I explain how I configured my Arch Linux machine, and control my desktop environment with XMonad.
With all the quantum-computing topics I lately hear, read, and talk about, I thought it’s high time to refresh some basics of statistical theory. Naturally, being a mathematical discipline, a great deal of formulas and equations are used to describe the nature of random outcomes. However, I prefer a more experimental approach where tests are executed and visualized with computer programs, which is why you’ll find Python code interlaced in this post. I encourage everybody reading this, to open a code editor, and play around with these programs.
Recently, I received note that IBM has several quantum processors which are available to the public for free. As computing and programming is a passion of mine and I like to try new things, I decided to give them a try. As anticipated, programming on such a chip is no easy task, because the underlying physics is completely different to conventional processors. So, here is my attempt of connecting the physics of the quantum realm with computing…
I’ve recently submitted the final version of my doctoral thesis to the Turicensia archive of the Zentralbibliothek Zurich. I still have a lot of information in my head about gravitational lensing which wants to get out. So, this post acts as a release to me and an informational blurp to you. I discuss the basics of gravitational lensing in a historical context, supported with a few mathematical excursions.
On October 29 2020 in midst of the COVID-19 pandemic, I defended my PhD thesis ‘Exploring models of lensing galaxies: On bridging the gap between observations, models, and simulations’ during the PhD colloquium. In order to make the virtual presentation easier to follow, I used the reveal.js framework to turn my presentation slides into a HTML/javascript website.
This post is a summary of a talk I was asked to give in front of my fellow PhD students. It basically was a tutorial on shell scripting and dotfile integration. I gave a live demo in which I demonstrated how I install my dotfile repository on a new machine. Dotfiles set defaults for many different programs, and make personalization of these programs easy to back up.
This post is a summary of a talk I was recently asked to give in front of my fellow PhD students. It basically was a git tutorial with a few extension from GitHub. I gave a live demo in which I demonstrated how I use git in my all-day work. Git really is a something worth knowing, and easy to learn (even if it sometimes doesn’t seem like it). It takes some time to pick it up, but the advantages outweigh the expenditure of time hundredfold, trust me!
With all the advances in rocket science and the talk about sending people to Mars in reuseable rockets, there is another method that could send people into space. The space elevator is a theoretical construct that could provide a much cheaper way to transport people or cargo into Earth’s orbit. In theory, it could even act as a construction base and launch pad for rockets to facilitate long space missions. All this sounds almost too good to be true… and it probably is. Even though one already has quite a good idea how this space elevator should look like, there are many hurdles to overcome. In this post I want to talk about the physical basics that make the space elevator possible.
Ever asked yourself why mountains are as high as they are? A short google search reveals that almost all of Earth’s highest mountains are located in the Himalaya around 8000 to 8500 meters, Mount Everest being the most famous and ascended. There is a mountain almost twice as tall though, Olympus Mons. Only thing is… it’s not a Terrestrial mountain. Olympus Mons resides on Mars, and as far as we know, it is with 21 km the tallest mountain in the Solar System. And it’s not the only mountain taller than Mount Everest, Mars has at least 4 mountains higher than 8 km. So why can mountains grow taller on Mars than on Earth? As yours might be, my first thought is gravity! Let’s try to estimate the maximal height of mountains on Earth, considering the only thing they have to fight against is gravity…
In the previous post I talked about Jekyll, a blog-aware, static website generator. In this posts I will talk about how I used it to create this website.
A few months ago, a friend asked me what web framework I use for my website and how I would decrease the latency inbetween database calls. After a short pause and a questioning facial expression from my part, he grinningly asked if I knew what he meant and if I even had a personal website. Although it’s not that of an uncommon state to be in, website-less that is, I unfortunately had to admit my shortcomings… but already in that moment I decided to build one from scratch.