Thoughts on AI, research, and the future of autonomous systems
As AI systems become more prevalent in safety-critical applications, the need for provably safe reinforcement learning algorithms becomes paramount. In this post, I explore the current challenges and future directions in safe RL research...
Read more →Traditional reinforcement learning assumes additive reward structures, but many real-world problems exhibit submodular properties. This post delves into our recent work on bridging submodular optimization with RL...
Read more →During my time with AMZ Racing at ETH Zurich, we built championship-winning autonomous race cars. This experience taught me valuable lessons about real-world AI deployment, sensor fusion, and the importance of robust systems...
Read more →I'm currently working on several blog posts about my research and experiences. Stay tuned for more insights into AI safety, reinforcement learning, and autonomous systems.
In the meantime, feel free to check out my publications or get in touch!