01
About
I am a machine learning engineer at Reducto, where I help train our vision-language models for document intelligence.
Previously, I was at Liquid AI, helping train the multimodal LFM2 and LFM2.5 family of models optimized for edge deployment. I've also collaborated with Reticular on improving sparse autoencoder (SAE) feature consistency.
I completed my M.Eng. at MIT CSAIL, where I was affiliated with the Distributed Robotics Lab and advised by Daniela Rus. I obtained my B.S. in Computer Science and Engineering from MIT.
02
Selected Papers
- Enforcing Orderedness to Improve Feature Consistency
Sophie L. Wang*, *, Nithin Parsan, John J. Yang
NeurIPS Mech Interp Workshop · 2025
- Continuous Autoregressive Generation with Mixture of Gaussians
, Tsun-Hsuan Wang, Ramin Hasani, Mathias Lechner, Alexander Amini
ICML Efficient Systems for Foundation Models Workshop · 2025
- LFM2 Technical Report
Liquid AI Team
arXiv · 2025
- Flex: End-to-End Text-Instructed Visual Navigation with Foundation Models
Makram Chahine, , Alaa Maalouf, Tsun-Hsuan Wang, Daniela Rus
IEEE Robotics and Automation Letters · 2026
- Gaussian Splatting to Real World Flight Navigation Transfer with Liquid Networks
*, Makram Chahine*, Alexander Amini, Ramin Hasani, Daniela Rus
Conference on Robot Learning · 2024
- Out of Distribution Generalization via Interventional Style Transfer in Single-Cell Microscopy
Wolfgang M. Pernice, Michael Doron, , Aditya Pratapa, Sultan Kenjeyev, Nicholas De Veaux, Michio Hirano, Juan C. Caicedo
CVPR Workshops · 2023
* Equal contribution