Ashutosh Adhikari

I work on cool stuff as an Applied Scientist at Microsoft Turing in Montreal.

Before that, I graduated with an MMath in Computer Science from the University of Waterloo where I worked on natural language processing and machine learning under the supervision of Jimmy Lin and Pascal Poupart. I had recieved the David R. Cheriton Graduate Scholarship at the University of Waterloo.

During my time at UWaterloo, I was a research intern at Mila from the fall of 2019 to the summer of 2020, where I worked with William L. Hamilton in collaboration with Microsoft Research, Montreal on using structured representations to develop generalizable policies for playing text-based games.

Before that, I completed my BTech (Honors) in Information and Communication Technology with minor in Computational Science in 2018 from DA-IICT, Gandhinagar, India. I was a research intern at Nanyang Technological University, Singapore in the winter of 2018 and spent the summer of 2017 at the Video Analytics Lab at the Indian Institute of Science, Bengaluru.

Email  /  Resume  /  Biography  /  Google Scholar

profile photo
Research

I'm broadly interested in natural language processing, graph representation learning and reinforcement learning. I have worked on simplifying models for document processing and developing general policies to solve text-based games.

fast-texture Learning Dynamic Belief Graphs to Generalize on Text-based Games
Ashutosh Adhikari*, Xingdi Yuan*, Marc-Alexandre Côté*, Mikulás Zelinka, Marc-Antoine Rondeau, Romain Laroche, Pascal Poupart, Jian Tang, Adam Trischler, William L. Hamilton
Accepted as a Poster at NeurIPS 2020
fast-texture Exploring the Limits of Simple Learners in Knowledge Distillation for Document Classification with DocBERT
Ashutosh Adhikari, Achyudh Ram, Raphael Tang, William L. Hamilton, Jimmy Lin
Accepted at the RepL4NLP workshop at ACL 2020
fast-texture DocBERT : BERT for Document Classification
Ashutosh Adhikari, Achyudh Ram, Raphael Tang, Jimmy Lin
Posted on arXiv.org
fast-texture A POMDP-based Context Aware Approach for Trust Modeling
Ashutosh Adhikari, Robin Cohen
Accepted at the TRUST workshop at AAMAS 2019
fast-texture Rethinking Complex Neural Network Architectures for Document Classification
Ashutosh Adhikari*, Achyudh Ram*, Raphael Tang, Jimmy Lin
Accepted as an Oral at NAACL-HLT 2019
fast-texture FLOPs as a Direct Optimization Objective for Learning Sparse Neural Networks
Raphael Tang, Ashutosh Adhikari, Jimmy Lin
Accepted at the Complex Deep Neural Networks with industrial applications workshop at NeurIPS 2018
fast-texture Comprehensive study of features for subject-independent emotion recognition
Ashutosh Adhikari, Savitha Ramasamy, Suresh Sundaram
Accepted as a Poster at IJCNN 2017

Website template taken from Jon Barron.