Ameet Deshpande

NLP Ph.D. Candidate at Princeton University
Researcher at Allen Institute for AI (AI2)


I am a computer scientist pursuing my Ph.D. at Princeton University, advised by Dr. Karthik Narasimhan. My research interests are broadly in the field of natural language processing (NLP), with special emphasis on sub-topics of AI safety, data-efficient algorithms, efficient NLP architectures, and multilingual NLP.

I am currently visiting Allen Institute for AI (AI2) and working with Dr. Ashwin Kalyan as part of the Aristo team.

I am deeply passionate about teaching and view disseminating knowledge to be as important as the research I do.

Previously, I received my B.Tech from IIT Madras, where I worked on reinforcement learning with Prof. Balaraman Ravindran and on NLP with Prof. Mitesh M. Khapra and Prof. Sutanu Chakraborti.

In my free time, I enjoy playing football (soccer), tennis, and the piano.


Dec 10, 2023 Was invited to be an oral session chair in EMNLP 2023 at Singapore.
Oct 9, 2023 Three papers accepted at EMNLP! Toxicity in ChatGPT, Conditional STS, and MUX-PLMs.
Oct 1, 2023 Our workshop titled PERSONALIZE: Personalization of Generative AI Systems will be co-organized with EACL 2024 in the beautiful and historic Malta.
Sep 24, 2023 Our work Toxicity in ChatGPT was featured on Wall Street Journal, along with a quote from my conversation with them.
Apr 25, 2023 Our work SemSup-XC is accepted at ICML 2023. SemSup-XC achieves state-of-the-art results on extreme classification.
Apr 12, 2023 Our work on AI safety which analyzes toxicity in ChatGPT was featured on media outlets like CNBC, VentureBeat, TechCrunch, and Gizmodo. A website summarizing our work can be found here.
Apr 12, 2023 Our work on AI safety titled “Toxicity in ChatGPT: Analyzing Persona-assigned Language Models” is out. Find it here. Work done with my wonderful collaborators at Princeton, AI2, and Georgia Tech.
Feb 20, 2023 Our pre-print titled “MUX-PLMs: Pre-training Language Models with Data Multiplexing” is now available! Find it here.
Feb 20, 2023 Our pre-print titled “SemSup-XC: Semantic Supervision for Zero and Few-shot Extreme Classification” is now available! Find it here.
Feb 15, 2023 I conducted a workshop at Princeton University on ChatGPT.