I am a computer scientist and a researcher commited to pursuing reliable AI.

I love exploring new areas, challenging old conventions, innovating novel solutions, and telling exciting stories. I strive to do meaningful work that contributes toward creating a good future for ourselves and our next generations...

Projects

I am interested in all kinds of topics that enrich our lives such as arts, sports, culture, and science. Throughout my studies and work, I have always been guided by my fascination with these. The projects below illustrate my interests more specifically.

AHA!

Art Human & AI

Through exhibitions at Science Gallery London and King's College London, I explore in collaboration with artist Alice White, the value of art as a mirror into the furthest reaches of the human mind and its implications for developing robust and generalizable AI.

See project

Safe and Trusted AI

PhD at King's and Imperial College

I joined the Safe and Trusted AI initiative by King's College London and Imperial College London as a PhD research student in the fall of 2020. So far, I have explored how to interpret fine-grained natural language understanding (NLU) models at individual prediction level as well as at more holistic level using novel visualization and explanation techniques. Currently, I am interested in how to incorporate causal modeling to better predict the behaviour of an NLU model in deployment.

Research profiles: KCL profile, Google Scholar, Semantic Scholar

See project

Golden Score

Human body pose optimization

Martial art is a part of me, my family, and my culture. It boosts my life with self-discipline, health, and excitement. In this project, under the supervision of Dr. Rita Borgo, I pay homage to the beauty of judo by aiming to 3D-reconstruct a judo scene from a single image. This project was awarded the Alan Fairbourn Memorial Prize for the Most Meritorious Final Year Project by King's College London and produced an academic paper, which was published at EuroGraphics 2022.

See project

Waterfall

Mobile app for iOS and Android

Learning is hard but everyone has to do it. That is why I care about making learning easier and more fun for everyone. The Waterfall technique is a time-efficient memorization method for learners, which allows you to minimize the time spent on easy items so you can focus more on the difficult ones. Traditionally, it involves the tedious task of preparing decks of flashcards, which is not possible to do without the right kit and environment, or buying decks, which are hardly ever tailored to your needs. This app allows you to create and use your decks digitally anytime anywhere and also to use decks created by other users.

See project

Contact me

Email: munkhtulga.battogtokh@kcl.ac.uk

© Munkhtulga Battogtokh