Arindam Mitra
Crafting next-generation foundation models that are not only powerful and trustworthy but also customizable and efficient, catering to a diverse range of existing and emerging applications. My work has been integrated into Bing and Science Engine, the latter of which empowers scientists across multiple enterprises with cutting-edge AI products.
Email  /  Google Scholar  / 
Semantic Scholar  /  Linkedin  / 
Twitter
|
|
Research
I am Arindam, a tiny timid soul born amidst another tiny timid one named Oodlabari, a village that I call home. Deep fog amuses me and probably that's why I find myself immersed in the boundless realm of artificial intelligence.
My zeal for natural language understanding has driven me to explore how machines can acquire knowledge and effectively
reason with it, forming the core of my research. Previously, I developed hybrid systems that integrated machine learning with formal representation and reasoning. However, the impressive capabilities of LLMs and neural models to
store knowledge within their weights and reason using additional contextual information have led me to question the necessity of separate tools for formal reasoning. My recent research efforts involve building large foundation
models that excel at learning from instructions, can follow detailed guidelines, and reason with utmost precision. I consider myself fortunate to be working on the problems I am passionate about at Microsoft Research, where I am
part of an exceptional team that shares my enthusiasm for natural language understanding and large-scale models.
|
|
Learning From Instructions/Prompting
|
|
Orca: Progressive Learning from Complex Explanation Traces of GPT-4
Subhabrata Mukherjee*, Arindam Mitra*, Ganesh Jawahar, Sahaj Agarwal, Hamid Palangi, Ahmed Awadallah
Arxiv
|
|
Instruction Tuned Models are Quick Learners
Himanshu Gupta, Saurabh Arjun Sawant, Swaroop Mishra, Mutsumi Nakamura, Arindam Mitra, Santosh Mashetty, Chitta Baral
Arxiv
|
|
Neural Reasoning
|
|
Improving biomedical information retrieval with neural retrievers
Man Luo, Arindam Mitra, Tejas Gokhale, Chitta Baral
AAAI Conference on Artificial Intelligence, 2022
|
|
Commonsense Reasoning with Implicit Knowledge in Natural Language
Pratyay Banerjee, Swaroop Mishra*, Kuntal Kumar Pal*, Arindam Mitra*, Chitta Baral
Automated Knowledge Base Construction (AKBC), 2021
|
|
Deeply embedded knowledge representation & reasoning for natural language question answering: A practitioner's perspective
Arindam Mitra, Sanjay Narayana, Chitta Baral
Workshop on Structured Prediction for NLP, 2020
|
|
Enhancing natural language inference using new and expanded training data sets and new learning models
Arindam Mitra, Ishan Shrivastava, Chitta Baral
AAAI Conference on Artificial Intelligence, 2020
|
|
Careful selection of knowledge to solve open book question answering
Pratyay Banerjee, Kuntal Kumar Pal*, Arindam Mitra*, Chitta Baral
Annual Meeting of the Association for Computational Linguistics (ACL), 2019
|
|
Hybrid Reasoning
|
|
Declarative Question Answering over Knowledge Bases Containing Natural Language Text with Answer Set Programming
Arindam Mitra, Peter Clark, Oyvind Tafjord, Chitta Baral
Proceedings of the AAAI Conference on Artificial Intelligence, 2019
|
|
Combining knowledge hunting and neural language models to solve the Winograd schema challenge
Ashok Prakash, Arpit Sharma, Arindam Mitra, Chitta Baral
Annual Meeting of the Association for Computational Linguistics (ACL), 2019
|
|
A generate-validate approach to answering questions about qualitative relationships
Arindam Mitra, Chitta Baral, Aurgho Bhattacharjee, , Ishan Shrivastava
Arxiv, 2019
|
|
Learning To Use Formulas To Solve Simple Arithmetic Problems
Arindam Mitra, Chitta Baral
Annual Meeting of the Association for Computational Linguistics (ACL), 2016
|
|
Learning to automatically solve logic grid puzzles
Arindam Mitra, Chitta Baral
Empirical Methods in Natural Language Processing (EMNLP), 2015
|
|
The NL2KR platform for building natural language translation systems
Nguyen Vo Arindam Mitra, Chitta Baral,
Annual Meeting of the Association for Computational Linguistics (ACL), 2015
|
|
Inductive Learning
|
|
Incremental and iterative learning of answer set programs from mutually distinct examples
Arindam Mitra, Chitta Baral
Theory and Practice of Logic Programming, 2018 (Jounral Publication)
|
|
Addressing a question answering challenge by combining statistical methods with inductive rule learning and reasoning
Arindam Mitra, Chitta Baral
AAAI Conference on Artificial Intelligence, 2016
|
|
Datasets
|
|
NumGLUE: A Suite of Fundamental yet Challenging Mathematical Reasoning Tasks
Swaroop Mishra, Arindam Mitra, Neeraj Varshney, Bhavdeep Sachdeva, Peter Clark, Chitta Baral, Ashwin Kalyan
Annual Meeting of the Association for Computational Linguistics (ACL), 2022
|
|
Enhancing natural language inference using new and expanded training data sets and new learning models
Arindam Mitra, Ishan Shrivastava, Chitta Baral
AAAI Conference on Artificial Intelligence, 2020
|
|
Declarative Question Answering over Knowledge Bases Containing Natural Language Text with Answer Set Programming
Arindam Mitra, Peter Clark, Oyvind Tafjord, Chitta Baral
AAAI Conference on Artificial Intelligence, 2019
|
Template from the awesome Jon
|
|