Abishai Ebenezer

Abishai Ebenezer

CS PhD Student at University of California, San Diego

About

I am a Computer Science PhD student at UC San Diego, advised by Dr. Trey Ideker. I previously was a Staff Research Scientist at the Institute of Protein Design, University of Washington, advised by Dr. David Baker. At the Baker Lab, I worked on building machine learning models that use diffusion and flow-matching algorithms to design T-Cell Receptors.

Prior to that, I worked as a Research Fellow at Microsoft Research, PROSE, where I built language models to fix formulae in Microsoft Excel and contributed to Visual Studio Copilot.

Outside of work, I love playing football, swimming, reading and writing blogs. I also am a huge Arsenal, Seahawks and RCB fan!

News

Papers

2024
FLAME: A Small Language Model for Spreadsheet Formulae. Harshit Joshi, Abishai Ebenezer, José Cambronero, Sumit Gulwani, Aditya Kanade, Vu Le, Ivan Radiek, Gust Verbruggen. AAAI Conference on Artificial Intelligence (AAAI 2024). PDF
2022
A Question-Centric Evaluation of Descriptive Answers Using Attention-Based Architecture. Abishai Ebenezer, et al.. IEEE International Conference on Confluence 2022. PDF
An Atypical Metaheuristic Approach to Recognize an Optimal Architecture of a Neural Network. Abishai Ebenezer, Arti Arya. International Conference on Agents and Artificial Intelligence (ICAART 2022). PDF