Medium

Alex Punnen

Pinned

In

Towards AI

by

Alex Punnen

·

Feb 14, 2025

Explaining Transformers as Simple as Possible through a Small Language Model

And understanding Vector Transformations and Vectorizations

Explaining Transformers as Simple as Possible through a Small Language Model

Pinned

In

Better ML

by

Alex Punnen

·

Sep 4, 2023

Exploring Large Language Models

Language Understanding, Task Generalization, Causal Reasoning, Explainability/Hallucinations, Quantisation, Fine-Tuning via Instruct…

Exploring Large Language Models

Pinned

In

Better Software

by

Alex Punnen

·

Apr 25, 2019

Optimizing any TensorFlow model using TensorFlow Transform Tools and using TensorRT

Model Optimization and reducing precision from FP32 to FP 16 for speedup and reducing graph size.

Optimizing any TensorFlow model using TensorFlow Transform Tools and using TensorRT

Pinned

Alex Punnen

·

Feb 26, 2022

Backpropagation Deep Dive

Back Propagation with Softmax output and CrossEntropy Loss

Backpropagation Deep Dive

Pinned

In

Better Software

by

Alex Punnen

·

Jul 4, 2022

From Data Warehouse to Data Lake to Data Lakehouse

What’s for what, what you need, and what are the advantages and limitations

From Data Warehouse to Data Lake to Data Lakehouse

In

Towards AI

by

Alex Punnen

·

1d ago

The AI SDLC Manifesto: Rethinking AI native Software Development

How Specification-Driven Development and machine-consumable knowledge can reshape software delivery

The AI SDLC Manifesto: Rethinking AI native Software Development

In

alexpunnen

by

Alex Punnen

·

Jun 18

A Sunday Morning Walk

More spiritual than the Sunday Church

A Sunday Morning Walk

In

Towards AI

by

Alex Punnen

·

May 15

Why Specification-Driven Development (SDD) is Not a Silver Bullet for AI-Assisted SDLC

Using a popular SDD GitHub Spec Kit and a real-world feature experiment

Why Specification-Driven Development (SDD) is Not a Silver Bullet for AI-Assisted SDLC

In

Towards AI

by

Alex Punnen

·

Apr 20

The RAG Problem

Why RAG looks Easy, but is really much Harder

The RAG Problem

In

Towards AI

by

Alex Punnen

·

Apr 12

Fine Tuning an LLM is Educational But not Very Useful Still for Knowledge Ingestion

Experience with Gemma 4 and other models

Alex Punnen

Alex Punnen

Distinguished Member of Technical Staff with 25 years building production software. Author of The Maths of Deep Learning. https://alexcpn.github.io/

Following

Help

Status

About

Careers

Press

Blog

Store

Privacy

Rules

Terms

Text to speech