Awarded the National Science and Technology (NST) Fellowship by the Ministry of Science and Technology, Bangladesh
Date:
I am honored to have been awarded the National Science and Technology (NST) Fellowship by the Ministry of Science and Technology, Bangladesh for my research project:
“VLM-Driven Multi-Step Inverse Models for Learning Agent-Centric Representations in RL”
I secured Merit Rank 11th out of 1260 MS/MSc candidates (Physical Sciences group) in this highly competitive selection process. 🔗 Official Announcement
🏆 About the NST Fellowship
The National Science and Technology (NST) Fellowship is a prestigious research grant awarded by the Ministry of Science and Technology (MoST), Bangladesh. It aims to support young researchers in advancing scientific innovation and contributing to their respective fields. The fellowship is highly competitive, recognizing outstanding proposals from undergraduate, master’s, and PhD researchers across various scientific domains.
📌 Research Project Overview
Title: VLM-Driven Multi-Step Inverse Models for Learning Agent-Centric Representations
🔍 Research Focus
This project explores how Vision-Language Model (VLM) representations can enhance agent control in RL through multi-step inverse dynamics models. The key focus areas include:
Leveraging VLM representations to extract semantic, high-level features relevant to control tasks.
Utilizing multi-step inverse models to learn control-relevant latent representations for RL agents.
Evaluating the approach in RL environments such as Matterport, Habitat, and Softgym.
This fellowship provides critical support for my ongoing research, and I look forward to sharing more insights as the project progresses!
A huge thank you to my advisors, mentors, and collaborators for their invaluable guidance and support throughout this journey. Looking forward to pushing the boundaries of AI & RL research!