I am Maxence Faldor, currently a research scientist intern at Sakana AI. I am pursuing a PhD at Imperial College London in the Adaptive and Intelligent Robotics lab. I am interested in 🧠 Artificial Intelligence ✨ Open-endedness 🦾 Robotics 🦎 Emergence. My research focuses on building open-ended and self-improving systems.

I am passionate about building intelligent systems that can learn, discover, and create in an open-ended fashion. Below are some of my projects in this area.

	OMNI-EPIC: Open-endedness via Models of human Notions of Interestingness with Environments Programmed in Code
	The AI CUDA Engineer: Agentic CUDA kernel discovery, optimization and composition
	Discovering Quality-Diversity Algorithms via Meta-Black-Box Optimization
	Toward Artificial Open-Ended Evolution within Lenia using Quality-Diversity
	CAX: Cellular Automata Accelerated in JAX

I am one of the organizers of ICARL (Imperial College Autonomous Reasoning & Learning), a student-led initiative at Imperial College London. Our goal is to foster a vibrant community for students and researchers interested in the future of AI.

Our main activities include:

• 🎥 Seminar Series: We invite leading researchers from academia and industry (sponsored by Google DeepMind, InstaDeep, and Iconic) to share their work. Past talks are uploaded to our YouTube channel.
• 📚 Reading Group: We meet bi-weekly to discuss recent publications in Deep Reinforcement Learning, Generative AI, and Autonomous Agents. The group is open to everyone—feel free to get in touch!

Follow our activities and join our community through our official channels: