Hi! My name is Thanmay Jayakumar, and I am an AI Resident at AI4Bhārat, Indian Institute of Technology Madras, India.

Broadly, my research interests are related to investigating the cross-lingual capabilities of Large Language Models (LLMs) and extending their support to low-resource languages, and I am guided by Dr. Anoop Kunchukuttan, Dr. Raj Dabre and Prof. Mitesh Khapra. I aspire to alleviate language barriers by improving tools for multilingual communication. I also work on building open-source datasets and models for Indian languages. Previously, I worked on Audio Retrieval during my undergraduate summer internship at IIT Kanpur, supervised by Prof. Vipul Arora.

I completed my Bachelors of Technology in Electronics & Communication Engineering at Visvesvaraya National Institute of Technology (VNIT) Nagpur. I developed my interest in Deep Learning and Natural Language Processing (NLP) at IvLabs, the Robotics & AI Lab of VNIT Nagpur, where I reviewed and implemented state-of-the-art architectures. At IvLabs, I also served as a mentor for junior undergraduates interested in pursuing NLP research. During my final year, I worked under Prof. Mansi Radke on Open Information Extraction and Prof. Anamika Singh on Image Captioning, the latter of which was for my bachelors thesis.

Apart from NLP, linguistics also captivates me. I have done significant study of various languages and their morphology, such as German, Persian, Indonesian, Chinese and several Indic languages. Please do say hi in your langauage and maybe we can have a chat in your native tongue. :)

CV / Resume

Email ID: [firstname][lastname]@gmail.com

Updates

Aug 2024:In Bangkok to attend ACL 2024. Excited to learn and connect with the NLP community!
Apr 2024:Our work, RomanSetu: Efficiently unlocking multilingual capabilities of Large Language Models models via Romanization. has been accepted at ACL 2024 Main Conference!
Feb 2024:Delighted to announce that our undergraduate work has been accepted at LREC-COLING 2024 Main Conference! Grateful to Prof. Mansi Radke for her continuous support and guidance.
Jan 2024:Announcing our first Hindi instruction-tuned model Airavata. Do read the technical report for more details!
Dec 2023:I'm in Singapore to attend EMNLP 2023. Please do attend my presentation at the NLLP Workshop!
Oct 2023:Our paper titled Large Language Models are legal but they are not: Making the case for a powerful LegalLLM has been accepted at Natural Legal Language Processing Workshop at EMNLP 2023! This is my first workshop paper at a *ACL venue.
Sep 2023:Delighted to join AI4Bharat as an AI Resident.
Apr 2023:Our paper, Attending to Transforms: A Survey on Transformer-based Image Captioning has been accepted at PCEMS 2023!
Aug 2022:Started work on Open Information Extraction, supervised by Prof. Mansi Radke, VNIT Nagpur.
Aug 2022:Started work on Automatic Image Captioning, supervised by Prof. Anamika Singh, VNIT Nagpur.
Jul 2022:Accepted into the IIIT-H's Advanced Summer School on NLP at Hyderabad, India. Project guided by Saumitra Yadav and Prof. Manish Shrivastava. Check out the Project Presentation.
May-Aug 2022:Accepted into the prestigious SURGE internship program at Indian Institute of Technology, Kanpur, India. Project on Spoken Term Detection (Audio Retrieval), supervised by Prof. Vipul Arora. Check out my Project Report.
Jun 2021:Started work on low-resource Neural Machine Translation at IvLabs, Visvesvaraya National Institute of Technology. Check out the Presentation.
May-Jul 2020:Started my Summer Internship at IvLabs, Visvesvaraya National Institute of Technology. Project on Automatic Speaker Recognition, supervised by Prof. Shital Chiddarwar.