PhD Dissertation Defense - Shushan Arakelyan

Title: 
Building Generalizable Language Models for Code Processing
 
Abstract:
Successful deployment of any AI model requires generalization to previously unseen, real-world scenarios. Lack of generalization in models can lead to outcomes ranging from reduced performance to potential legal liabilities. In this thesis, I explore generalization challenges in large language models for code processing. I will discuss three different generalization concerns that language models for code processing can exhibit, and present my progress in building models that can overcome those. 1) I will start by discussing compositional generalization issues, where models must adapt to previously unseen instruction combinations 2) Next I will talk about out-of-domain generalization, and how distribution shifts within single projects or corporations can affect model performance, and how to overcome it. 3) Finally, I will talk about generalization of advanced models to programming languages with fewer resources.
 
Venue: SAL 213
 
Date/Time: August 23, 1pm-3pm
 
Names of the Dissertation Defense Committee members: 
Xiang Ren (chair), Morteza Dehghani, Aram Galstyan, Mukund Raghothaman
 

Location: Henry Salvatori Computer Science Center (SAL) - 213

Audiences: Everyone Is Invited

Contact: Ellecia Williams