Pulkit Verma

Postdoc at Interactive Robotics Group, MIT

pulkit_verma.png

I am a Postdoctoral Associate in the Interactive Robotics Group at CSAIL and AeroAstro, MIT, working with Prof. Julie Shah. I completed my Ph.D. from the School of Computing and Augmented Intelligence at Arizona State University, working with Prof. Siddharth Srivastava. During my Ph.D., I was also affiliated with UC Berkeley’s Center for Human-Compatible Artificial Intelligence, and ASU’s Center for Human, Artificial Intelligence, and Robot Teaming. I have completed M.Tech. from Department of Computer Science & Engineering at IIT Guwahati, advised by Prof. Pradip K. Das.

My primary interests lie in AI safety, AI planning, action-model learning, interpretability, and analysis of abstractions. My research focuses on safe and reliable behavior of AI agents. I am currently investigating the minimal set of requirements in an AI system that would enable a user to assess and understand the limits of its safe operability.

In the past, I have worked in the areas of bio-inspired robotics, speech processing, and internet of things.


News

Sep 05, 2025 Our work PDDL-Instruct: Enhancing Symbolic Planning Capabilities in LLMs through Logical Chain-of-Thought Instruction Tuning has been accepted at ICAPS 2025 Workshop on Planning in the Era of LLMs.
Sep 05, 2025 Our work A Collaborative Numeric Task Planning Framework based on Constraint Translations using LLMs has been accepted at ICAPS 2025 Workshops on Human-Aware and Explainable Planning and Planning in the Era of LLMs.
Aug 03, 2025 Serving as the Workflow Chair for ICAPS 2026.
Jul 22, 2025 Our work An LLM-powered Collaborative Task Planning Framework has been accepted at ICAPS 2025 in the System Demonstrations and Exhibits Track. More details to follow soon!!
Jun 24, 2025 Giving a talk on Interpretable Model Learning for Taskable AI Systems at IEEE-ICAD 2025.

See older news...