Developed robust evaluation methods for dialogue systems using LLMs and human-based assessments. Organised the SCI-CHAT workshop at EACL 2024 and published its findings on simulating conversational intelligence.
Project Link →Worked on 'Reducing Bias in Language Models using Deep Reinforcement Learning'. Formulated the problem as a multi-armed bandit environment to assess and reduce bias in Question Answering models.
Project Link →Formulated a genre classification problem to distinguish fiction vs non-fiction. Developed 'Fictometer' web app and 'NewsChase' mobile app to detect manipulative content.
Project Link →Studied Nash Equilibrium using graph theory and simplicial complexes. Investigated dynamic games and strategic bargaining (Rubinstein Model). Defended MS thesis in Game Theory.
Project Link →Explored stochastic processes like Markov Chains and Pólya’s Urn Model under Prof. Manjunath Krishnapur.