publications

2025

1. VRB

   Vision Reward Bench: Towards Scalable Benchmarking Without Supervision

   Alexi Gladstone^*, Siddhartha Mishra^*, and Heng Ji

   2025

   Manuscript in Preparation for ICML 2025

2024

1. EBWM

   

   Cognitively Inspired Energy-Based World Models

   Alexi Gladstone, Ganesh Nanduru, Md Mofijul Islam, and 3 more authors

   arXiv preprint arXiv:2406.08862, 2024

   Under Review at ICML 2025

   arXiv Bib Blog

   @article{gladstone2024cognitively,
     title = {Cognitively Inspired Energy-Based World Models},
     author = {Gladstone, Alexi and Nanduru, Ganesh and Islam, Md Mofijul and Chadha, Aman and Li, Jundong and Iqbal, Tariq},
     journal = {arXiv preprint arXiv:2406.08862},
     year = {2024},
   }

2. Embodied Referring Expression Comprehension Through Multimodal Residual Learning

   Md Mofijul Islam, Alexi Gladstone, Sujan Sarker, and 4 more authors

   2024

   Under Review at ICLR 2025

2023

1. EQA-MX

   

   Eqa-mx: Embodied question answering using multimodal expression

   Md Mofijul Islam, Alexi Gladstone, Riashat Islam, and 1 more author

   In The Twelfth International Conference on Learning Representations, 2023

   Spotlight Acceptance Rate 5%

   Bib HTML

   @inproceedings{islam2023eqa,
     title = {Eqa-mx: Embodied question answering using multimodal expression},
     author = {Islam, Md Mofijul and Gladstone, Alexi and Islam, Riashat and Iqbal, Tariq},
     booktitle = {The Twelfth International Conference on Learning Representations},
     year = {2023},
   }

2. Patron

   

   Patron: perspective-aware multitask model for referring expression grounding using embodied multimodal cues

   Md Mofijul Islam, Alexi Gladstone, and Tariq Iqbal

   In Proceedings of the AAAI Conference on Artificial Intelligence, 2023

   Bib HTML

   @inproceedings{islam2023patron,
     title = {Patron: perspective-aware multitask model for referring expression grounding using embodied multimodal cues},
     author = {Islam, Md Mofijul and Gladstone, Alexi and Iqbal, Tariq},
     booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
     volume = {37},
     number = {1},
     pages = {971--979},
     year = {2023},
   }

2022

1. CAESAR

   

   CAESAR: An embodied simulator for generating multimodal referring expression datasets

   Md Mofijul Islam, Reza Mirzaiee, Alexi Gladstone, and 2 more authors

   Advances in Neural Information Processing Systems, NeurIPS 2022 Datasets and Benchmarks Track , 2022

   Bib HTML

   @article{islam2022caesar,
     title = {CAESAR: An embodied simulator for generating multimodal referring expression datasets},
     author = {Islam, Md Mofijul and Mirzaiee, Reza and Gladstone, Alexi and Green, Haley and Iqbal, Tariq},
     journal = {Advances in Neural Information Processing Systems, },
     volume = {35},
     pages = {21001--21015},
     year = {2022},
   }