https://kauuu.github.io/topics/personal/Recent content in Personal on Kaushik's PortfolioHugoen-usMon, 06 Apr 2026 13:30:00 +0200https://kauuu.github.io/posts/my-new-blogging-setup/Mon, 06 Apr 2026 13:30:00 +0200https://kauuu.github.io/posts/my-new-blogging-setup/<p>Hello everyone,</p> <p>Over the easter weekend, I came across a tweet that Obsidian were looking for a new engineer to add to their team, and it was just their 4th engineer. This was insane, given how big their app is. This made me consider switching to <em>Obsidian</em> as my primary note-taking app (over <em>Notion</em>). However, I needed more to be convinced, and that is when I stumbled upon the <em>file over app</em> philosophy<sup id="fnref:1"><a href="#fn:1" class="footnote-ref" role="doc-noteref">1</a></sup> proposed by the CEO of Obsidian. What it says is the following:</p>https://kauuu.github.io/posts/my-first-blog/Sat, 01 Mar 2025 15:45:04 +0100https://kauuu.github.io/posts/my-first-blog/<p>Does this work?</p>https://kauuu.github.io/about/Mon, 01 Jan 0001 00:00:00 +0000https://kauuu.github.io/about/<p>Hey there &#x1f44b;</p> <p>I&rsquo;m Kaushik — currently pursuing an MSc in Computer Science at ETH Zürich 🇨🇭, where I&rsquo;m focusing on understanding how AI systems truly work. Prior to this, I completed my BSc in Computer Science and Engineering at TU Delft 🇳🇱.</p> <p>My current interests revolve around <strong>multimodal language models</strong> — how they process the information, whether they truly reason, and what <em>interpretability</em> can teach us. I believe understanding the internals of these models is crucial in <a href="https://safe.ai/ai-risk">designing systems that are more controllable, reducing catastrophic risks</a>. I&rsquo;m actively looking to explore this space further. If you have any interesting projects or ideas in this area, feel free to reach out!</p>