Bridging Algorithms and Biology | AI-Driven Innovator & Quantum-Curious Developer
I’m Dinmay Kumar Brahma — crafting intelligent systems and scientific tools at the intersection of artificial intelligence, quantum computing, and biotechnology. Currently studying at IIT Guwahati and developing projects that push the frontier of computation and cognition.
A collection of projects showcasing my work and skills.
This is a Cuda applied ML Library so that anyone can use GPU Powered ML with Ease in Python.
This is a from scratch implementation of Meta's Free Transformers paper.
Flash attention implementation from scratch.
PINNs can successfully incorporate physical laws to predict system behavior even when training data contains noise.
No description available.
No description available.
Transformers from scratch implemented GQA,RoPE,RMS-Norm and trained on that code
Universal Transformers Decoder Only architecture
This is transformers from scratch coded with py and torch, kv cache and GQA(Grouped Query Attention) is implemented.
No description available.
It was a great learning.....
I've written this code to understand the K-Means from the deep 😅
No description available.
AiSQL: AI-Powered SQL Query Generator
Adaptive-Offset-Based-Quantization-Method
This is ESRGAN_Model Finetuned with 4k video frames.
No description available.
Lox is a lightweight, dynamically-typed programming language with a simple and minimal syntax.
REPO (Context Re-Positioning) is a research paper by Sakana AI that addresses challenges in long-context language models through a novel approach to context handling.
Sharing knowledge about AI, machine learning, and software development through articles and tutorials.
The AI model that doesn’t just write code — it actually sees what it’s building. Let me be honest with you. For the past six months, I’ve been watching the “AI...
Inside the MoE architecture, Native Vision, and the swarm capabilities that just broke the benchmarks. Look, I still remember where I was sitting when GPT-4 dro...
Read on Medium
This article discusses observed behavior from an early, leaked checkpoint. Features and performance may change in the final release. Conceptual visualization of...
Read on Medium
A technical analysis of the coding-first architecture, conditional memory, and the "Two Jobs" problem. For the last two years, the “scaling law” has been a brut...
Read on Medium
Why the future of game-playing AI isn’t “train a bot per title,” but “pretrain once, adapt everywhere.” The Hook: The “Status Quo” Frustration Have you ever wat...
Read on Medium
Why the future of AI isn’t generating JPEGs, but generating Photoshop files. If you’ve used Stable Diffusion, Midjourney, or Flux, you know the frustration. You...
Read on MediumMy work centers on building intelligent systems while exploring how computation can model, optimize, and eventually emulate complex natural processes. I have hands-on experience with Python-based AI development, algorithmic problem-solving, and experimental system design. Beyond implementation, I am deeply interested in the theoretical foundations of computation, including quantization methods, optimization theory, and numerical modeling. I enjoy working close to first principles—understanding why a system works before scaling how it works. My long-term research interests span artificial cognition, biotechnology-inspired computation, and quantum mechanics, where I aim to contribute original ideas rather than incremental variations. I approach projects as evolving experiments—designed to test assumptions, uncover structure, and push technical boundaries.
Bsc(hons) Data science & Artificial intelligence
Have a project in mind or want to collaborate? I'd love to hear from you. Drop me a message and I'll get back to you as soon as possible.
Kolkata, West Bengal
