🧠 AI Research Paper Notes

A structured, in-depth repository documenting my journey through landmark Artificial Intelligence research papers.

📖 Introduction

Welcome to my AI Research Notebook! This repository serves as a systematic collection of my reading notes, analyses, and insights on foundational and state-of-the-art research papers in Artificial Intelligence. It is designed to track my progress and demonstrate my ability to comprehend, synthesize, and document complex AI research.

🎯 Why I Created This Repository

I created this repository to showcase my research journey and paper-reading skills while preparing for AI Research Internships at institutions like OIST, RIKEN, Max Planck, ETH Zurich, EPFL, KAUST, and other prominent research labs.

💡 Learning Objective

My core objective is to deeply understand the architectures, mathematical foundations, and intuitive insights behind breakthroughs in Deep Learning, Computer Vision, and Natural Language Processing, moving beyond superficial knowledge to develop a rigorous research mindset.

🏛️ Repository Philosophy

This repository is maintained like a serious research notebook:

• Consistency: Regular, structured documentation of newly read papers.
• Depth of Understanding: Focusing on why and how things work, not just what they do.
• Long-term Learning: A scalable knowledge base that will grow over the years.
• Research Mindset: Highlighting observations, analogies, open questions, and critical analysis.
• Clean Documentation: Organized, minimal, and highly readable notes.

---

🗺️ Research Paper Roadmap

📚 Papers by Domain

👁️ Computer Vision

• AlexNet
• ResNet
• Vision Transformer (ViT)
• Segment Anything (SAM)
• DINO

🗣️ Natural Language Processing & Foundational Models

• Attention Is All You Need
• BERT
• GPT (Original)

🚀 Large Language Models & Efficient Tuning

• LoRA
• QLoRA
• Llama 3 Technical Report
• DeepSeek-V3 Technical Report

🔍 Multimodal & Retrieval

• CLIP
• Retrieval-Augmented Generation (RAG)
• REALM

⚙️ Core Architectures & Optimization

• Batch Normalization

---

📊 Progress Tracker

#	Paper	Domain	Status	Notes
01	Attention Is All You Need	NLP / Architecture	✅ Read	Notes
02	AlexNet	Computer Vision	⬜ Unread	-
03	ResNet	Computer Vision	⬜ Unread	-
04	Batch Normalization	Optimization	⬜ Unread	-
05	Vision Transformer (ViT)	Computer Vision	⬜ Unread	-
06	CLIP	Multimodal	⬜ Unread	-
07	BERT	NLP	⬜ Unread	-
08	GPT (Original)	NLP	⬜ Unread	-
09	LoRA	LLM / Fine-Tuning	⬜ Unread	-
10	QLoRA	LLM / Fine-Tuning	⬜ Unread	-
11	Retrieval-Augmented Generation (RAG)	LLM / Retrieval	⬜ Unread	-
12	REALM	LLM / Retrieval	⬜ Unread	-
13	Segment Anything (SAM)	Computer Vision	⬜ Unread	-
14	DINO	Computer Vision	⬜ Unread	-
15	Llama 3 Technical Report	LLM	⬜ Unread	-
16	DeepSeek-V3 Technical Report	LLM	⬜ Unread	-

---

📂 Folder Structure

AI-Research-Paper-Notes/
│
├── README.md
│
├── 01-Attention-Is-All-You-Need/
├── 02-AlexNet/
├── 03-ResNet/
├── 04-Batch-Normalization/
├── 05-Vision-Transformer/
├── 06-CLIP/
├── 07-BERT/
├── 08-GPT/
├── 09-LoRA/
├── 10-QLoRA/
├── 11-RAG/
├── 12-REALM/
├── 13-Segment-Anything/
├── 14-DINO/
├── 15-Llama-3/
└── 16-DeepSeek-V3/

Inside Every Paper Folder

[Paper-Name]/
├── README.md        # Polished final understanding and summary
├── My-Notes.md      # Raw thoughts, observations, analogies, and questions
├── Resources.md     # Links to the paper, official code, blogs, and lectures
└── assets/          # Diagrams, handwritten notes, and self-made illustrations

---

🔄 Documentation Workflow

For every paper I read, I follow this structured workflow:

1. Initial Pass: Skim the abstract, introduction, and conclusion. Jot down initial thoughts in My-Notes.md.
2. Deep Dive: Read the methodology and experiments carefully. Draw architectures and add them to assets/.
3. Synthesis: Write a clean, structured summary in the folder's README.md.
4. Reference Gathering: Collect the best explanations, code, and videos in Resources.md.

---

📈 Repository Statistics

• Total Papers Documented: 1
• Total Domains Covered: 5
• Latest Paper Read: Attention Is All You Need

---

🔮 Future Papers

List of papers to be added in the next phases of learning.

• To be updated...

---

📚 References

• arXiv
• Papers with Code
• Hugging Face Papers