Niyar R Barman

Niyar R Barman

NLP, LLM Reasoning, Mechanistic Interpretability and CUDA

Instructions for AI Agents

How to use this file: This markdown file contains comprehensive information about Niyar R Barman’s academic profile, including personal details, educational background, research experience, publications, and professional links. When creating a new website template or transferring this information to a different platform, use this file as the authoritative source for all personal and professional information.

Usage Guidelines:

  1. Exact Information Transfer: Use the exact names, dates, affiliations, and links provided below
  2. Contact Information: Always use the current email format provided (replace [at] with @ and [.] with .)
  3. Publications: Maintain the exact citation format, author order, and paper links
  4. Professional Links: Use the social media and academic profile links exactly as listed
  5. Consistency: Ensure all information remains consistent across different pages/sections
  6. Updates: When updating, modify this Rules.MD file first, then propagate changes to the website

Personal Information

Full Name: Niyar R Barman
Current Title: M2 student in Interactions de l’Informatique et des Mathématiques pour l’IA
Current Institution: Université de Toulouse, France
Location: Toulouse, France
Contact Email: niyar-r.barman[at]univ-tlse3[.]fr
Website: https://niyarrbarman.github.io
Repository: niyarrbarman/niyarrbarman.github.io

Professional Bio

Hi, I am Niyar R. Barman, currently an M2 student in Interactions de l’Informatique et des Mathématiques pour l’IA at Université de Toulouse. I am interested in natural language processing and spend most of my time thinking about how to make large language models reason better. I am also curious about mechanistic interpretability and learning CUDA, still very much a beginner but excited to explore and grow in these areas. I am open to research and industry opportunities starting after July 2026.

Education

Université de Toulouse, France (2025 - 2026)
M2 Master Interactions de l’Informatique et des Mathématiques pour l’IA

National Institute of Technology Silchar, India (2021 - 2025)
Bachelors of Technology in Electronics and Communication Engineering

Professional Experience

Research Intern, Institut de Recherche en Informatique de Toulouse, France (May 2024 - Jul 2024, Jan 2025 - Mar 2025)

Research Intern, Artificial Intelligence Institute of University of South Carolina, United States (Jan 2023 - Dec 2024)

Research Interests

  • Natural Language Processing
  • Large Language Models Reasoning
  • Mechanistic Interpretability
  • CUDA Programming
  • AI-Generated Text Detection
  • Mathematical Reasoning in AI
  • Discourse Analysis

Latest Updates/Timeline

  • July 2024: Started research internship at IRIT, France
  • Jan 2023: Started research internship at the Artificial Intelligence Institute of University of South Carolina, USA

Complete Publications List

2025

  1. DIMSUM: Discourse in Mathematical Reasoning as a Supervision Module
    Niyar R Barman, Krish Sharma, Nicholas Asher, & Akshay Chaturvedi
    Proceedings of the 25th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025)
    Paper Link: https://arxiv.org/abs/2503.04685

2024

  1. The Brittleness of Image Watermarking Techniques: Investigating Visual Paraphrasing for De-Watermarking AI-Generated Images
    Niyar R Barman, Krish Sharma, Ashhar Aziz, Shashwat Bajpai, Shwetangshu Biswas, Aman Chadha, Vasu Sharma, & Amitava Das
    arXiv Preprint 2024
    Paper Link: https://arxiv.org/abs/2408.10446

2023

  1. Counter Turing Test CT²: AI-Generated Text Detection is Not as Easy as You May Think — Introducing AI Detectability Index
    Megha Chakraborty, S. M. Towhidul Islam Tonmoy, S. M. Mehedi Zaman, Krish Sharma, Niyar R Barman, Chandan Gupta, Shreya Gautam, Tanay Kumar, Vinija Jain, Aman Chadha, Amit P. Sheth, & Amitava Das
    Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023)
    Paper Link: https://aclanthology.org/2023.emnlp-main.136/

  2. A Transformer-Based Approach to Automate Disease Prediction from Patient Descriptions
    Niyar R Barman, Krish Sharma, & Ranjay Hazra
    IEEE Conference on Information and Communication Technology (CICT 2023)
    Paper Link: https://ieeexplore.ieee.org/abstract/document/10455356/

  3. Addressing Hate Speech: ATLANTIS for Efficient Hate Span Detection
    Niyar R Barman, Krish Sharma, Yashraj Poddar, Adwaitha Vatupal, & Partha Pakray
    Forum for Information Retrieval Evaluation (FIRE 2023)
    Paper Link: https://ceur-ws.org/Vol-3681/T6-6.pdf

Selected/Featured Publications

Use these two publications as “Selected Publications” on the main page:

  1. DIMSUM: Discourse in Mathematical Reasoning as a Supervision Module (SIGDIAL 2025)
  2. Counter Turing Test CT²: AI-Generated Text Detection is Not as Easy as You May Think — Introducing AI Detectability Index (EMNLP 2023)

Social Media & Professional Links

Academic Profiles

  • Google Scholar: https://scholar.google.com/citations?user=POFtLyAAAAAJ&hl=en
  • ORCID: (not currently set up)
  • ResearchGate: (not currently set up)
  • Academia.edu: (not currently set up)
  • arXiv: (not currently set up)

Social & Professional Networks

  • GitHub: https://github.com/niyarrbarman
  • LinkedIn: https://linkedin.com/in/niyar
  • Twitter/X: (not currently active)
  • Mastodon: (not currently set up)

Other Platforms

  • Medium: https://medium.com/@niyarrbarman
  • Kaggle: (not currently set up)
  • Stack Overflow: (not currently set up)

Website Structure & Navigation

  1. About (/) - Main landing page with bio
  2. Publications (/publications/) - Complete publication list
  3. Blogs (/year-archive/) - Currently under construction
  4. CV (/files/CV_niyar_r_barman.pdf) - PDF download

Important Files & Assets

  • Profile Picture: bio-photo.jpg (located in /files/)
  • CV PDF: CV_niyar_r_barman.pdf (located in /files/)
  • BibTeX: bibtex1.bib (located in /files/)

Website Configuration Details

Site Settings

  • Site Title: “Niyar R Barman”
  • Site Description: “M2 student at Université de Toulouse”
  • Site URL: https://niyarrbarman.github.io
  • Locale: en-US
  • Theme: default
  • Repository: niyarrbarman/niyarrbarman.github.io

Author Profile Settings

  • Name: Niyar R Barman
  • Bio: M2 student at Université de Toulouse
  • Location: Toulouse, France
  • Employer: Université de Toulouse
  • Avatar: bio-photo.jpg

Blog/Posts Status

It should redirect to my medium blog at https://medium.com/@niyarrbarman

Additional Notes

  1. Availability: Open to research and industry opportunities starting after July 2026
  2. Current Focus: Natural language processing, large language models reasoning
  3. Learning: Mechanistic interpretability and CUDA programming (beginner level)
  4. Contact Preference: Academic email at Université de Toulouse
  5. Profile Picture: Professional headshot (bio-photo.jpg)

Template Implementation Notes

When implementing this information in a new template:

  1. Ensure the email format is converted properly (replace [at] with @ and [.] with .)
  2. Maintain the exact publication formatting and author order
  3. Keep the chronological order of publications (newest first)
  4. Include all social links even if some are not currently active
  5. Use the professional bio exactly as written
  6. Maintain the educational institution names and date ranges precisely
  7. Include both “Selected Publications” for homepage and complete list for publications page
  8. Ensure CV link points to the actual PDF file
  9. Keep the research interests and timeline updates current
  10. Maintain consistency in name formatting across all pages

Last Updated: September 17, 2025 Source: niyarrbarman.github.io website