Arnau Marin-Llobet

Hi there, thanks for vising my internet space.

My research is generously supported by fellowships from the Rafael del Pino Foundation, NSF NeuroPAC, and RCC at Harvard.

I’m a PhD student at Harvard broadly fascinated by the study of natural and artificial black boxes (natural brains and AI models) and how we can better understand and leverage their inner workings.

Recently, my interests have centered on using AI agents to automate, accelerate, and augment scientific discovery, as well as exploring mechanistic interpretability in large-scale models.

Selected recent projects

I created SpikeAgent, an AI agent based on large language models (LLMs) that does spike sorting and neural signal processing for you.

I also developed BCI-Agent, a multi-agent system that can infer cell-type specific brain-computer interfaces across different species—without the need for transgenic animals.

If you’d like to learn more, feel free to contact me or any team member through GitHub or email!

Collaborations

I’ve had the privilege of working with exceptional researchers including Prof. Yulia Sandamirskaya at the Neuromorphic Computing Lab at Intel and Zhaw, Prof. Mavi Sanchez-Vives at ICREA and University of Barcelona, and Prof. Benjamin Sanchez-Terrones at University of Utah; among others.

I’m always excited to collaborate at the intersection of AI, neuroscience, and engineering - feel free to reach out!

Personal

When I’m not in the lab, I’m dangerously good at making handmade pizza.