I am currently a Postdoctoral Associate in the Mens, Manus, and Machina team at the Singapore-MIT Alliance for Research and Technology Centre, where I work with Daniela Rus, Armando Solar-Lezama, and Bryan Low.
Prior to this, I was a Research Fellow in Department of Computer Science at National University of Singapore, where I worked with Bryan Low. I obtained my Doctor of Philosophy (Ph.D.) degree from Indian Institute of Technology Bombay (IIT Bombay), where Manjesh K. Hanawal and N. Hemachandra supervised me. My doctoral thesis received two awards: Naik and Rastogi Excellence in Ph.D. Thesis Award and COMSNETS Best Ph.D. Thesis Award.
My research focuses on developing adaptive and efficient AI systems that autonomously make complex decisions in dynamic real-world environments.
Arun Verma is a Postdoctoral Associate in the Mens, Manus, and Machina team at the Singapore-MIT Alliance for Research and Technology Centre, where he works with Daniela Rus, Armando Solar-Lezama, and Bryan Low.
Prior to this, he was a Research Fellow in Department of Computer Science at National University of Singapore, where he worked with Bryan Low. He obtained his Ph.D. degree from Indian Institute of Technology Bombay (IIT Bombay), where Manjesh K. Hanawal and N. Hemachandra supervised him. His doctoral thesis received two awards: Naik and Rastogi Excellence in Ph.D. Thesis Award and COMSNETS Best Ph.D. Thesis Award.
His research focuses on developing adaptive and efficient AI systems that autonomously make complex decisions in dynamic real-world environments.
- Sep 2026 BarrierSteer and Contextual Flow Matching accepted at NeurIPS 2026; SFT–RL Budget Allocation accepted at NeurIPS 2026 Workshop.
- Sep 2026 Invited to serve as an Area Chair (AC) for ICLR 2027.
- Aug 2026 Tutorial From Heuristics to Guarantees: Bandit and RL Blueprint for Reliable Agentic-AI accepted at NeurIPS 2026.
- Aug 2026 SFT–RL Budget Allocation (Main) and Prompt Optimization with Human Feedback (Findings) accepted at EMNLP 2026.
- Jul 2026 MMT: Immediate Exact Federated Unlearning accepted at Machine Learning Journal (ECML PKDD 2026 Journal Track).
