Sashank Neupane

Undergraduate Student

Deep Learning
Computer Vision
ML Security
I am a current undergraduate student at New York University Abu Dhabi. I am interested in deep learning, particularly in the field of computer vision. My current research focuses on ML security where I am exploring different kinds of backdoor attacks on vision models and how to defend against them.
Sashank Neupane

B.S. in Computer Science

New York University Abu Dhabi

Abu Dhabi, UAE

Aug 2021 - Current

High School

St. Xavier's College Maitighar

Kathmandu, Nepal

June 2017 - July 2019


Experience

Research Assistant

NYU Abu Dhabi

eBrains Lab
Abu Dhabi, UAE

I work on ML Security, focusing on backdoor attacks on vision models.

May 2022 - Current

Reference Assistant

NYU Abu Dhabi

Library
Abu Dhabi, UAE

I work a part-time job at the library as a reference assistant and help patrons with their research needs.

Aug 2022 - Current

Teaching Assistant

NYU Abu Dhabi

Computer Engineering
Abu Dhabi, UAE

I was a TA for the Computer Programming for Engineering course taught by Professor Eid in Fall 22. I helped students with lab work and held office hours.

Aug 2022 - Dec 2022


Projects

Blog It image

Blog It

JavaScript
Handlebars
MongoDB
TailwindCSS

The blogsite is a platform for people to write and share blogs to the world. The site will support all creation, update, and deletion of existing/new blogs from the site itself. The site also supports user authentication.

MySQL LMS image

MySQL LMS

Python
MySQL
Flask

MySQL Database Interface for a Learning Management System. The project is a simple interface for a Learning Management System that allows users to interact with a MySQL database.

Backdoor Toolbox image

Backdoor Toolbox

Python
PyTorch
TorchVision

A toolbox for backdoor attacks on vision models. The toolbox provides a simple interface to create backdoor attacks on vision models and test them.

BackdoorDocs image

BackdoorDocs

JavaScript
React
Next.js

A documentation site for the Backdoor Toolbox. The site provides detailed documentation on how to use the toolbox and its features.