Alexandra is a PhD candidate in the Department of Mechanical and Aerospace Engineering at Princeton University. Her research explores the safeguarding of robotics and AI against dual-use dilemmas. Originally from Moscow, Russia, she earned her Bachelor of Science in Mechanical Engineering from the California Institute of Technology (Caltech). Alexandra’s previous research, conducted in Princeton's IRoM lab, centered on developing generalization algorithms for machine learning as applied to robotics. Alexandra received the Gordon Y. S. Wu Fellowship and the Henry Ford II Award for academic excellence. She is also actively involved in leadership roles, previously serving as Vice President of the Princeton Graduate Student Government and as a coordinator for the Princeton Robotics Seminar series.
Research Interests
Motivated by the increased weaponization of robotics, Alexandra’s research focuses on developing ethical and safe autonomous systems, with a particular interest in mitigating the dual-use risks of robots and AI. Inspired by Asimov’s First Law of Robotics, Alexandra’s current project explores the concept of an “Asimov Box”—a hardware and software system designed to empower robots to act safely, ethically, and autonomously. This research aims to integrate real-time risk assessment frameworks, enabling robots to recognize and prevent actions that could cause harm or misuse, especially in sensitive or high-stakes environments. By employing innovative techniques like neural network pruning and adversary monitoring algorithms, we enable robots to “unlearn” harmful behaviors and uphold ethical standards. Beyond technical design, together with collaborators from First Law Robotics Lab, Alexandra investigates the social implications of autonomous systems, focusing on enhancing public trust and reshaping perceptions of robots in society. Through interdisciplinary collaboration, this research aims to create responsible, non-weaponized robotics that foster a positive, constructive role for AI and robotics in diverse civilian applications, from public spaces to secure facilities.
Publications
Allen Z. Ren, Anushri Dixit, Alexandra Bodrova, Sumeet Singh, Stephen Tu, Noah Brown, Peng Xu, Leila Takayama, Fei Xia, Jake Varley, Zhenjia Xu, Dorsa Sadigh, Andy Zeng, and Anirudha Majumdar, Robots That Ask For Help: Uncertainty Alignment for Large Language Model Planners, Conference on Robot Learning (CoRL), 2023.
Talks
Good Robot, Bad Robot: Preventing Dual-Use in AI and Robotics, Princeton School on Science and Global Security, Princeton University, October 2024.