MEXT Research Student Scholar 2023

  University of Fukui, Fukui, Japan

  Email - adniajay@gmail.com

 neuro     notes     blog   sights   delialah

I am PhD student (Hasegawa Lab) in Intelligent Information Systems at University of Fukui, where I am studying why LLMs misbehave and what causes emergent misalignment in LLMs. I am currently being advised by Prof. Tatsuhito Hasegawa.

Previously, I worked on - Causal Reinforcement Learning in Causal Set Theory, as a Master's student in Information Science, advised by Dr. Tomoyuki Yamakami from Dept. of Information Sciences, at Yamakami Lab, University of Fukui, Japan.

As a undergraduate student, I worked in the field of misinformation analysis and social media intelligence under the guidance of Dr. Basant Agarwal (Dept. of CS&E, IIIT Kota), Dr. Priyanka Harjule (Dept. of Mathematics, MNIT Jaipur), and Dr. Azizur Rahman (Dept. of CSE and Mathematics, Charles Sturt University). Our research primarily focused on exploring the game theoretic aspects of misinformation sharing on social media, and investigation of the epidemiological characteristics of misinformation spread. My work in the same has been consolidated in the form of published papers, patents and small-scale replicable projects/prototypes. I am also a member of the India-Slovenia Bilateral Scientific and Technological Cooperation project funded by DST, India on "Cross-lingual knowledge transfer for social media analysis in less-resourced languages during COVID-19"

I was an active member of the e-LDA Lab at IIIT Kota (e-Learning and Data Analytics Lab) supervised by Dr. Basant Agarwal, and Dr. Priyanka Harjule. Our research was primarily on the applications of explainable AI for misinformation detection and mitigation. Our work on Red Fort riots 2021, "Explainable AI-augmented misinformation monitoring on social media during crisis events" was one of the founding works of the e-LDA Lab.

Apart from Computational Complexity and Misinformation analysis, I've also done research in biostatistics, psychology and philosphy out of sheer interest and respect for the fields. I've published works in all these respective fields. My first internship was as an "Biostatistics Data Scientist" at Softsensor AI Labs based out of Texas.

Besides academic roles, I've volunteered my time as a freelance translator for a bootstrapped art-startup in Poland, DailyArt. I've also designed multiple multi-lingual social awareness posters as a part of various community awareness programs led by ISRC in support for public health intiatives. I've also been a student volunteer at NeurIPS 2020 & 2021, ICML 2021, NAACL 2021 and ICON 2019 (IIIT Hyderabad).

Turing Machines Behind the Horizon: Modeling Black Hole Interiors as Transfinite Limited Turing Machines
Ajay Agarwal
Soft Computing: Theory and Applications (SoCTA), 2022
Paper

Variable tradeoff between diversification and volatility on risk-averse approaches to crypto-allocation for Indian retail investors
Ajay Agarwal
International Conference On Business Research And Innovation, 2022
Paper

The Accidental Checkmate - Understanding the intent behind sharing misinformation on social media during COVID-19 using game theory
Ajay Agarwal, Basant Agarwal, Priyanka Harjule, Azizur Rahmna, and VN Manjunath Aradhya
Journal of Experimental & Theoretical Artificial Intelligence
Paper

Understanding the role of Feature Engineering in Fake News Detection
Ajay Agarwal, Basant Agarwal, and Priyanka Harjule
Soft Computing: Theory and Applications (SoCTA), 2021
Paper