Francisco M. López
PhD Candidate in Developmental AI Frankfurt Institute for Advanced Studies
My research aims to understand how infants can autonomously develop sensorimotor control and coordination, using a combination of intrinsically-motivated reinforcement learning, unsupervised reprensentations, and hierarchical architectures. I have also developed several open-source research platforms, including MIMo, BabyBench, and DREAM.
In my spare time, I enjoy creating Strands games based on papers I read or talks I attend. You can play them here.
In my spare time, I enjoy creating Strands games based on papers I read or talks I attend. You can play them here.
Updates (scroll for more):
- May 2025: Our paper "Humans and variational autoencoders agree: attractive faces are average and feminine, not symmetric and masculine" was accepted at CCN 2025.
- April 2025: Presented our paper "Hierarchical Residuals Exploit Brain-Inspired Compositionality" at ESANN 2025 in Brugge, Belgium.
- May 2024: Presented our paper "Self-Calibrating Saccade-Vergence Interactions" at ICDL 2024 in Austin, Texas.
- November 2023: Presented our paper "Eye-Hand Coordination Develops from Active Multimodal Compression" at ICDL 2023 in Macau, China.
- June 2022: Our paper "MIMo: A Multimodal Infant Model for Studying Cognitive Development in Humans and AIs" was accepted at ICDL 2022.
- November 2021: Started my PhD with Prof. Dr. Jochen Triesch at the Frankfurt Institute for Advanced Studies, Germany.
- April 2021: Defended my Master's thesis, titled "A study on the spatial representation of numbers using context‑dependent associative memories", at the Faculty of Science, Universidad de la República, Uruguay.