PhD student, Northeastern University
I'm a PhD student at Northeastern, in the Data, Networks, and Algorithms Lab. I work on large language models: how to run them across several machines, including small ones, and how to change something they know without retraining the whole model.
A few things I'm working on right now.
This continues JARVIS, which splits a language model across several devices. We added a way for the remaining devices to cover for one that drops out. We tried it with Gemma-2B on 18 radios in the Colosseum testbed, and on seven Raspberry Pis.
Training a neural network so that, once it is split across machines, those machines have to send less to each other, without losing accuracy.
Looking for cheaper ways to correct one fact in a language model, without retraining it and without undoing everything else it already knows.
I studied data science in Barcelona. I'm now doing a PhD in Boston.
Northeastern University · Boston
Large language models, in the Data, Networks, and Algorithms Lab. I work on running them across several machines, and on changing what they know without retraining them.
Genesys Lab, Northeastern · Boston
My bachelor's thesis was on using deep learning to classify wireless signals. That's what got me into this.
Abi Global Health · Remote
I helped monitor their machine learning systems and worked with the lab on new models.
Universitat Politècnica de Catalunya · Barcelona
GPA 8.89/10. Honors in Discrete Math & Logic, Mathematical Optimization, Search & Analysis of Information, and Advanced Topics of Data Science.
Email is the best way to reach me.