Skip to content
View SushainDevi's full-sized avatar

Block or report SushainDevi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
SushainDevi/README.md

Typing SVG

🎓 Pursuing a Bachelor of Technology in Computer Science & Engineering at MIT WPU, Pune.

🧠 Active Researcher on project NIKA, focused on engineering Neuro-Symbolic architectures and advanced System-2 reasoning systems.

📚 Deeply exploring LLM Interpretability, Epistemic Agency, truth manifolds, and mechanistic steering mechanisms.

🖥️ Bridging neural models with symbolic logic processing frameworks while exploring foundational transformer implementations from scratch.

Research & Projects

Here are some of my highlighted pinned projects on GitHub:

Project Name Description GitHub Link
NIKA A core research initiative focused on Neuro-Symbolic architectures and System-2 reasoning capabilities. Investigating LLM interpretability paradigms, including epistemic agency dynamics, Möbius steering, and representations of truth manifolds within deep neural systems. GitHub
LLaMA-3-Like-Transformer-Model-Implementation The project demonstrates how to build a transformer-based large language model from scratch by incorporating components like tokenization, rotary positional encoding (RoPE), attention mechanisms, feedforward networks, and more. GitHub
DocChat DocChat is an AI-powered Document Summarization and Chatbot System designed to automate the extraction, cleaning, and summarization of text from various document formats. Built using state-of-the-art machine learning models, it provides a robust solution for interacting with uploaded documents through a user-friendly chatbot interface. GitHub
Music-Recomendation-SpotifyAPI This repository contains code for a music recommendation system that utilizes the Spotify API to extract music data and provide personalized recommendations based on user input. The system employs a hybrid approach, combining content-based filtering and popularity-based scoring to generate relevant recommendations. GitHub
TextToVideo-Model This project implements a text-to-video generation system using deep learning techniques. It creates videos based on textual prompts by generating sequences of images that depict moving shapes, specifically circles, in various directions and transformations. GitHub
OnlineLibrary-master Online library management system is a web application designed to facilitate the management of books. The frontend is built using HTML, CSS, and JavaScript, providing a user-friendly interface for users to interact with. GitHub

Tools & Technologies

🖥️ Programming Languages:

Python C++ SQL HTML CSS JavaScript

⚙️ Frameworks & Tools:

PyTorch TensorFlow Hugging Face Docker Streamlit

🌐 APIs & Focus Areas:

OpenAI API Gemini API Spotify API

🗄️ Databases:

SQLite MySQL

🔬 Research & Core Domains:

Neuro-Symbolic AI LLM Interpretability System-2 Reasoning Deep Learning NLP Generative AI

📚 Science & Support Libraries:

Torchvision SciPy OpenCV


Currently Coding & Listening to:
Spotify Logo

Pinned Loading

  1. Blender_Animation Blender_Animation Public

    a 30-second Blender animation where a garage opens to reveal a BMW car and a sports bike. During the animation, the BMW car exits the garage and runs along a path for about 3 seconds. This suggest…

  2. Music-Recomendation-SpotifyAPI Music-Recomendation-SpotifyAPI Public

    This repository contains code for a music recommendation system that utilizes the Spotify API to extract music data and provide personalized recommendations based on user input. The system employs …

    Python 1

  3. TextToVideo-Model TextToVideo-Model Public

    This project implements a text-to-video generation system using deep learning techniques. It creates videos based on textual prompts by generating sequences of images that depict moving shapes, spe…

    Python 1

  4. LLaMA-3-Like-Transformer-Model-Implementation LLaMA-3-Like-Transformer-Model-Implementation Public

    The project demonstrates how to build a transformer-based large language model from scratch by incorporating components like tokenization, rotary positional encoding (RoPE), attention mechanisms, f…

    Python 1

  5. DocChat DocChat Public

    DocChat is an AI-powered Document Summarization and Chatbot System designed to automate the extraction, cleaning, and summarization of text from various document formats. Built using state-of-the-a…

    Python 1

  6. OnlineLibrary-master OnlineLibrary-master Public

    online library management system is a web application designed to facilitate the management of books. The frontend is built using HTML, CSS, and JavaScript, providing a user-friendly interface for …

    CSS