CV

Postdoctoral Research Associate

Department of Informatics, King's College London (KCL)

Researcher in Natural Language Processing (NLP).

• Working on assessing and improving the robustness of NLP systems.
• Built adversarial evaluation frameworks to identify vulnerabilities in LLMs under Unicode-based stylistic and linguistic perturbations, uncovering systematic robustness failures across diverse subpopulations.
• Developed interpretability-based methods to enhance the robustness of LLMs.
• Designed red-teaming benchmark to evaluate safety robustness of open-source LLMs across multiple languages.
• Mentor in KCL Womxn in STEM.

Research Assistant

Department of Informatics, King's College London

Research Assistant in the Department of Informatics, King's College London.

• Researcher in Natural Language Processing

Research Fellow

Indian Institute of Technology Patna, India

Worked on fine-grained and coarse-grained sentiment analysis from text and multimodal content. Teaching Assistant for NLP | Python | Innovation Lab

• Researcher in Natural Language Processing

PhD

Indian Institute of Technology Patna

Natural Language Processing

Masters

National Institute of Technology Patna

Computer Science and Engineering

B.Tech.

Guru Nanak Dev Univ. RC Jalandhar

Computer Science and Engineering

Google Travel Award

Google India

To travel and present a paper at EMNLP.

Microsoft Travel Award

Microsoft India

To travel and present a paper at EMNLP.

ACM Travel Award

ACM India

To travel and present a paper at ECIR.

Diversity and Inclusion Grant

EMNLP

To travel and present a paper at EMNLP.

Department Gold Medalist (M.Tech.)

National Institute of Technology Patna (NITP)

Sentiment, Emotion, Sarcasm & Hate speech Classifier

This project is done in collaboration with CENTRE FOR DEVELOPMENT OF TELEMATICS (CDOT) Delhi. We developed a multilingual surveillance system capable of analyzing sentiment, emotion for English and Hindi languages, Sarcasm and Hate speech detection for English.

• Natural Language Processing

Information Diffusion in Online Social Networks

This project is implemented with deep learning approach in python language as a part of M.Tech thesis which predicts how far a piece of information will travel in the online social network.

• Natural Language Processing