# David Oniani

> David Oniani is an AI researcher and software engineer working as a Senior Applied Scientist on the Special Projects Team within Amazon Ads. His research focuses on Efficient AI Methods, AI Alignment, AI for Science, Large Language Models, and Software Systems.

This file serves as a directory of LLM-friendly content and pages on David Oniani's website (oniani.org).

## Bio

- [Biography](https://oniani.org/bio): Detailed biography, professional history, academic background, and general interests.

## Blog

- [Autoregressive Transformer](https://oniani.org/blog/transformer): Guide on the autoregressive transformer architecture, implementing a GPT-like Large Language Model from scratch in Python/PyTorch.
- [COVID-19 and Transition Into Full-Time Summer Work](https://oniani.org/blog/covid-19-and-transition-into-full-time-summer-work): Reflection on the challenges of transitioning to online learning and remote work during the early stages of the COVID-19 pandemic.
- [Mayo Clinic - First Impressions](https://oniani.org/blog/mayo-clinic-first-impressions): Initial impressions and experiences doing AI research at the Mayo Clinic.
- [Privacy for the Web](https://oniani.org/blog/privacy-for-the-web): Practical recommendations and suggestions for improving web privacy.
- [The Essence of Programming - Functional Approach](https://oniani.org/blog/the-essence-of-programming-functional-approach): General overview and introduction to functional programming paradigms, type theory, and lazy evaluation.

## Clar

- [Compact Linear Algebra Reference (CLAR)](https://oniani.org/clar): A concise, quick-reference guide summarizing key linear algebra concepts.

## From Scratch

- [From Scratch](https://oniani.org/from-scratch): Walkthroughs and implementations of foundational machine learning algorithms and data structures built from the ground up.

## Links

- [Links](https://oniani.org/links): Curated directory of external links, research resources, profiles, and pages.

## Notes

- [Batch, Group, and Layer Normalization](https://oniani.org/notes/batch_group_and_layer_normalization): Notes comparing different normalization techniques in deep learning.
- [Building a Machine Learning Model](https://oniani.org/notes/building_a_machine_learning_model): Notes on the process, metrics, and steps involved in designing and constructing ML models.
- [Exploding and Vanishing Gradients](https://oniani.org/notes/exploding_and_vanishing_gradients): Notes discussing gradient scaling problems and solutions in deep neural networks.
- [Gradient Descent](https://oniani.org/notes/gradient_descent): Notes on gradient descent optimization algorithms, variants, and mathematical foundations.
- [Information Theory](https://oniani.org/notes/information_theory): Notes summarizing key information-theoretic metrics like entropy, mutual information, and divergence.
- [Learning to Rank](https://oniani.org/notes/learning_to_rank): Notes on learning-to-rank algorithms and formulations used in search and recommendation.
- [Model Performance and Compression](https://oniani.org/notes/model_performance_and_compression): Notes covering model quantization, pruning, and efficiency methods.
- [Norms](https://oniani.org/notes/norms): Notes on mathematical vector norms and their definitions.
- [Recommender Models](https://oniani.org/notes/recommender_models): Notes outlining collaborative filtering, matrix factorization, and deep recommendation methods.
- [Regularization](https://oniani.org/notes/regularization): Notes on regularization techniques (L1, L2, dropout) used to prevent model overfitting.
- [Pointer Types in Rust](https://oniani.org/notes/rust_ptr): Notes outlining box, reference, raw pointers, and smart pointer types in the Rust programming language.
- [Statistical Hypothesis Testing](https://oniani.org/notes/statistical_hypothesis_testing): Notes detailing hypothesis formulation, significance testing, p-values, and statistical power.

## Places

- [Places](https://oniani.org/places): List of countries, cities, and places David has visited or lived in.

## Research

- [Research](https://oniani.org/research): Academic research, publications, and scientific studies in AI/ML.

## Software

- [Software](https://oniani.org/software): Open-source projects, libraries, and tools designed and developed by David.

## Tools

- [Tools](https://oniani.org/tools): Hardware setup and software/tool stack recommendations.