Luca Scimeca

I received my BEng in Artificial Intelligence and Software Engineering from the University of Edinburgh in 2017 with First Class Honors, graduating top of my class and receiving the Howe Prize for the best performance in Artificial Intelligence.

At the University of Cambridge, I completed my PhD in Prof. Fumiya Iida's Biologically Inspired Robotics Laboratory, before joining Naver Corp. as a visiting AI research scientist, where I worked on representation learning, robustness, and generalization in deep learning.

Postdoctoral fellowships took me to Harvard University and Dana-Farber, where I used machine learning to answer fundamental questions in tumor biology and immunotherapy, and to Yoshua Bengio's group at the Mila AI Institute, where I focused on generative AI and probabilistic inference.

Most recently, at Sapient Intelligence, I applied the same ideas to reasoning, language, and vision, working on HRM-Text and Unigen, state-of-the-art models for their size.

Currently working on something new at the intersection of AI & Science, more soon!

I am interested in the interplay between foundational AI research — generative AI, reasoning, probabilistic inference, multimodal perception, and representation learning — and its applications to critical domains, as well as their downstream implications for fairness, explainability, and safe inference.

I am also enthusiastic about the impact of machine learning on science and engineering, and have been fortunate to work across vision, computational biology (deep RL, bio-sequence design), robotics (visual and tactile action/perception, multimodal time series), and even astrophysics (gravitational lensing and inverse problems).

For collaborations, reach me at luca.scimeca@live.com.