->Developed benchmarks and evaluation methodologies for code generation with LLMs, focusing on evaluation of non-functional requirements of the generated code and code comprehension.->Developed a one-shot federated learning solution.

Working on the ASIST project to model beliefs of the agents in a multi-agent setting for a search and rescue task with Prof. Katia Sycara

Worked on the Empirical Analysis of Privacy-Preserving Learning for the open-source library Vowpal Wabbit with the Real World Reinforcement Learning Team at Microsoft Research NYC.

Worked on efficient Train Scheduling using Reinforcement Learning.

Experimented with class activation maps in NLP to understand the words affecting the textual entailment prediction being made by the CNN model.