Hello, my name is Sonia.


About me

I care about understanding humans as thinking, feeling, computational machines and using these insights to build artificial intelligence that better serves a diversity of human intelligence.

 Currently, I’m interested in combining models and methodologies from cognitive science, human-computer interaction, and natural language processing to understand and improve LLMs’ abilities to be safe, flexible, and cooperative social partners.

Harvard School of Engineering & Applied Sciences

I am a third-year PhD student in Computer Science, where I am advised by Tomer Ullman and Elena Glassman. I am grateful to be supported by an NSF Graduate Research Fellowship and to be a part of the inaugural cohort of Kempner Institute Graduate Fellows at Harvard.

The Allen Institute for Artificial Intelligence

Previously, I was a Predoctoral Young Investigator on the Semantic Scholar team, working with Doug Downey and Daniel Weld on generating diverse descriptions of scientific concepts.

Princeton Computational Cognitive Science Lab

I’m also grateful to have worked with Tom Griffiths and Robert Hawkins on developing models of human word-color associations and studying the mental representations that enable us to flexibly communicate.

If you would like to chat about research, or are a woman/minority student considering graduate school in Psychology or Computer Science, please feel free to reach out to me at soniamurthy [at] g [dot] harvard [dot] edu.

Preprints

One fish, two fish, but not the whole sea: Alignment reduces language models’ conceptual diversity
Sonia K. Murthy, Tomer Ullman, Jennifer Hu
(under review)
[pdf]

Publications

Comparing the Evaluation and Production of Loophole Behavior in Humans and Large Language Models
Sonia K. Murthy, Kiera Parece, Sophie Bridgers, Peng Qian, Tomer Ullman
EMNLP Findings (2023)
[pdf]

an earlier iteration of this work appeared in the Proceedings of the First Workshop on Theory of Mind in Communicating Agents @ ICML 2023

ACCoRD: A Multi-Document Approach to Generating Diverse Descriptions of Scientific Concepts
Sonia K. Murthy, Kyle Lo, Daniel King, Chandra Bhagavatula, Bailey Kuehl, Sophie Johnson, Jon Borchardt, Daniel S. Weld, Tom Hope, Doug Downey
EMNLP System Demonstrations (2022)
[pdf][demo][dataset + code][AI2 blog feature]

an earlier iteration of this work appeared in the Proceedings of the Fifth Widening Natural Language Processing Workshop @ EMNLP 2021.

Response distribution of color-concept associations for the 10 concepts with lowest and highest variability.

Shades of confusion: Lexical uncertainty modulates ad hoc coordination in an interactive communication task
Sonia K. Murthy, Thomas L. Griffiths, Robert D. Hawkins
Cognition (2022)
[pdf][code]

Other work

Towards a computational model of human word-color associations
Sonia K. Murthy. Advised by Tom Griffiths.
Undergraduate thesis, Princeton University (2020)
[abstract][pdf]
NCWIT Collegiate Award Finalist (2020)
[press]

© 2024 Sonia Murthy