Published at the 24th ACM Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2025. Co-authored with Saptarashmi Bandyopadhyay, Thomas Goldstein, and David Jacobs.
Post-training Vision-Language-Action models using multi-agent rewards to incentivize safe driving behaviors and zero-shot coordination.
Applied Multi-Agent Reinforcement Learning methods to develop Recommender Systems for Explainable AI with integrated human feedback.
Trained models to selectively forget targeted datasets while preserving performance using LIME and interpretability frameworks.
