https://blog.tbeolet.com/Recent content on Tristan BeoletHugoen-usMon, 16 Oct 2023 00:00:00 +0000https://blog.tbeolet.com/posts/offline-rl-glycemia-control/Mon, 16 Oct 2023 00:00:00 +0000https://blog.tbeolet.com/posts/offline-rl-glycemia-control/My first paper: training offline RL agents on real closed-loop data for type 1 diabetes, and personalizing them per patient without a simulator.https://blog.tbeolet.com/about/Mon, 01 Jan 0001 00:00:00 +0000https://blog.tbeolet.com/about/<p>I&rsquo;m Tristan Beolet. I work on AI, robotics and reinforcement learning, and this blog is where I write up the things I build, break and fix along the way.</p> <p>Expect notes on training agents, getting policies to behave on real hardware, and the tooling in between. The code for project posts lives in separate repositories, linked from each post.</p> <ul> <li>GitHub: <a href="https://github.com/TristanBEOLET">TristanBEOLET</a></li> <li>LinkedIn: <a href="https://www.linkedin.com/in/tristan-beolet">tristan-beolet</a></li> </ul>