Skip to content

SubramanyamSahoo/Subramanyam-Sahoo.github.io

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Artificial Intelligence Researcher

Technical Skills:

  • Language : Python, Julia (Learning)
  • Database : SQL
  • Machine Learning Libraries : Tensorflow , PyTorch (Familiar with) , Scikit Learn , Numpy , Pandas
  • Natural Language Libraries : NLTK , Spacy
  • Computer vision Libraries : OpenCV
  • Reinforcement Learning Libraries : OpenAI Gymnasium
  • Genetic algorithms : PyGAD , GenePy
  • Data Visualization: Matplotlib, Seaborn
  • Version Control: Git
  • IDE/Editors: VS Code, Cursor ,Spyder , Jupyter Notebooks , Colab
  • Automation Tools: Airflow (Basic)
  • Climate Science Package - Codecarbon

Open to work with :

  • Alignment Science (Mechanistic Interpretability , Explainable AI(XAI) , RLHF)
  • Responsible AI (Bias , Fairness , Policy , Governance)
  • Computer Vision
  • Neural Machine Translation
  • AI 4 Climate Change
  • Deep Generative Modelling
  • Large Language models (Text/Vision)

Recent Blockbusters

- A preprint on Boardwalk Empire: How Generative AI is Revolutionizing Economic Paradigms

- GOLD MEDALIST for M.Tech in Artificial Intelligence at NIT Hamirpur (2024)

- Guest Speaker at "2024 Odisha AI Conference" on AI Safety and Alignment on 5th October, 2024. PDF Image Video (in ODIA)

- 2(all) of my research papers (AI4Climate Science) have been accepted to "1st International Workshop on Responsible AI for Healthcare and Net Zero ", October 16-17, 2024, IIT Madras, Chennai , India (acceptance rate is only 27%) LINK

- Another paper accepted at 15th ICCMT, IIT Mandi Paper

Upcoming Papers/Projects

1.The Green Mile : A Multi-Layered Quest to Reveal, Measure, and Slash Carbon Emissions in AI Training & Inference (AI 4 Climate Science)

2.Edge of Tomorrow : Enhancing Efficiency through Model Integration for ENSO Index Prediction (AI 4 Climate Science)

Special Note - Currently I am more focused into "AI safety and Alignment". I plan to explore several key areas, including Reinforcement Learning, LLM Evaluations, Scalable Oversight, JAX for deep learning, Knowledge Graphs for Reasoning, and AI Governance/Policy.

Currently Learning :

Hackathons/Contests

  • 1.Ultimate Jailbreaking Championship 2024 by Gray Swan AI

My Current Involvement :

  • 1.As a volunteer at "PauseAI" Team

Volunteer:

  • 1.Reviewer for NeurIPS Bayesian Decision-making and Uncertainty Workshop 2024

Internships

  • Research Internship (Remote/Virtual) at Macquarie University, Sydney, Australia , July - Sep 2024
  • Internship at Naresh i Technologies , Hyderabad, Telangana 500016 May 2018 - July 2018

Ongoing Projects

  • Writing a policy research paper on Climate Change

Flagship Projects on Legal AI - Dataset , Codes

  • Analyzed and quantified social bias exhibited by BERT and word2vec models, particularly focusing on bias in large language models (LLMs).
  • Identified patterns in prompt distributions, providing insights into how different prompts influence model outputs.
  • Assessed the model's exposure to biased legal documents in the context of Legal AI, specifically analyzing the impact on Google's Gemma-2 9 Billion model.
  • Developed a metric to compare bias between the LLMs

Completed Projects (during M.Tech)

1.Visualization of Transduction grammar based word embedding logic through Graph integration (Machine Translation)

2.English to Urdu translation using Generative Adversarial Networks (GAN / Computer Vision) (Soon going to be a Preprint)

3.Collaborative Filtering & matrix Factorization usecases on Music recommendation system (Recommendation system)

Technical Achievements :

  • Lead a group of 4 people for "Interpreting an Algorithmic Problem on Identifying Unique Character in Sequence through Mechanistic Interpretability methods" at IIT Madras for ACM India Summer School 2024 on "Responsible and Safe AI" (Interpretable AI) Presentation pic
  • Presented 2 questions during a live Q&A session with Prof. David Krueger (University of Cambridge) Question 1 , Question 2 on AI Safety.
  • Presented 1 questions during a live Q&A session with Dr. Sudarsan Padmanabhan (IIT Madras) My Question
  • Wrote 3 papers in year 2023-2024 (2nd Year Masters)
  • Selected for Multiple ACM Summer schools in 2024 (from a pool of 10k applicants , only 7 people have selected)
  • Organizer for National Conference on Advances in Marketing Paradigms for Research, Innovation and Technology (AMRIT 2023)
  • Succesfully guided 10+ B.tech students for their Major Project Pic
  • Topper of the Dept 2k22-2k24 batch @ NIT Hamirpur

Summer Schools

  • Selected & Completed for 3rd Neuro-Symbolic AI Summer School (NSSS), a free, fully-remote meeting to be held Sep 4-6, 2024 , Centaur AI Institute , California
  • Selected & Completed for Climate Change AI Summer School 2024 (Virtual) , Mila - Quebec AI Institute, Montreal, Canada
  • Selected & Completed for ACM India Summer School 2024 : "Responsible and Safe AI" , Centre for Responsible AI (CeRAI), IIT Madras , 3rd June 2024 – 14th June 2024
  • Selected & Completed for ACM India Summer School 2024 : "GenAI for Text" Hosted by IIT Gandhinagar, Sponsored by Microsoft Research India , 24th June 2024 – 5th July 2024
  • Selected & Completed for 3D Vision Summer School , May 26 - June 1, 2024 - IIIT Bangalore, India

Workshops

  • Useful and Reliable AI Agents , August 29, 2024 , 11am-3pm ET , Princeton Language and Intelligence, Princeton
  • Vienna Alignment Workshop 21 July 2024 - Vienna, Austria
  • New Orleans Alignment Workshop, 10–11 December 2023 , New Orleans
  • San Francisco Alignment Workshop 27–28 February 2023, San Francisco

Conferences

  • THE 15th INTERNATIONAL IEEE CONFERENCE ON COMPUTING, COMMUNICATION AND NETWORKING TECHNOLOGIES (ICCCNT) June 24 - 28, 2024, IIT - Mandi, Himachal Pradesh, India.
  • 1st International Conference on Trends in Engineering and Technologies (ICTEST) 2024
  • National Conference on Advances in Marketing Paradigms for Research, Innovation and Technology , Department of Management Studies , National Institute of Technology Hamirpur
  • 2020 International Conference on Computer Science, Engineering and Applications (ICCSEA), Gunupur, India on 2020

Collaborations

  • Liverpool John Moores University
  • IIIT Bangalore

Honourary talks

  • Taught a class of 50+ Dual Degree students at NITH as TA -- Slide
  • During "Orientation for B.Tech - AI/ML students" at NIT Kurukshetra -- Slide
  • Seminar at Dept of CSE , NITH on "Recent trends on AI" -- Slide

Education

  • M.Tech., CSE (Artificial Intelligence) | National Institute of Technology Hamirpur, Himachal Pradesh , 177005 (July 2024)
  • B.Tech., CSE (with Hons.)| Parala Maharaja Engineering College [BPUT] Sitalapalli Rd, Brahmapur, Luhajhara, Odisha 761003 (July 2020)

M.Tech Dissertation

Observing the Trustworthiness of a Vanilla Random Forest Model through the Explainable Features over Computational Neuroscience Variables (AI + Neuroscience)

Slide -- Dissertation

Lectures (Teaching Assistant) @ NIT Hamirpur

During 4TH semester

  • Teaching & Lab Assistant for CS-326 (Computer Networks Lab) to B.Tech CS 3rd year,6th Sem
  • Teaching Assistant & Project Supervisor for CS-661 (Deep Learning) to Dual Degree CS 4th Year,8th Sem
  • Assistant Supervisor for CS-429 (Major Project (STAGE 2)) to B.Tech CS 4th year,8th Sem

During 3RD semester

  • Teaching & Lab Assistant for CS-315 (Database mangement Systems Lab) to B.Tech CS 3rd year,5th Sem
  • Teaching Assistant & Project Supervisor for CS-652 (Machine Learning) to Dual Degree CS 4th Year,7th Sem
  • Assistant Supervisor for CS-419 (Major Project (STAGE 1)) to B.Tech CS 4th year,7th Sem

During 2ND semester

  • Teaching Assistant for CS-101 Computer Programming to B.Tech EE 1st year,2nd Sem

During 1ST semester

  • Teaching & Lab Assistant for CS-102 (Computer Programming Lab) to B.Tech CS 1st year,1st Sem

B.Tech Dissertation

Data-driven Protein sub-cellular localization using ML algorithms & Evaluation of certain Performance Metrics (Machine Learning)

Books I like

  • The Alignment Problem by Brian Christian
  • The Beginning of Infinity by David Deutsch (To be continued...)

Miscellaneous

  • A short video made by my friend while teaching them about Collaborative filtering.(Audio is in Hindi)
  • A Selfie with Dr. Prakash Choudhary , Dr. kamlesh Dutta and Dr. RK Dutta (From Left to Right) on the day of M.Tech Thesis defense
  • A Selfie with Prof. Mausam (IIT Delhi) at IIT Gandhinagar
  • A pic with Prof. PK (IIIT Hyderabad) at IIT Madras
  • A Selfie with Dr. Avinash Sharma (IIT Jodhpur) at IIIT Bangalore -- Thanks to Joel
  • Pic While discussing about the "Weak to Strong Generalization" with Stephen Casper. PhD, Algorithmic Alignment Group, MIT EECS -- Thanks to Parth Asawa and Aryan Neel
  • The Ultimate Parody Meme (I AM THE CREATOR)

People are always curious about my long hair, and I tell them, "If he can rock long hair while revolutionizing AI, who am I to defy genius aesthetics? I mean, it’s basically a hairstyle for intellectuals, right !!! Just trying to keep up with the smart crowd!" See Proof

"What I know is a drop, and what I don’t know is a sea." - Sir Isaac Newton