RSSAmplifier

Blog

David An

Recent content on David An

davidan.devRSS feed ↗48 posts

Latest posts

Fine Tuning Llama 3.2B with Unsloth

In this article, we will be fine tuning the Llama 3.2B model with Unsloth on the Spider 1.0 SQL dataset. The goal of the article is to improve the SQL capabilities of a general Llama 3.2B model. Prerequisites Before we get started, we assume that the reader has access to a GPU which they are able to use for training. Additionally, we assume that the reader has a Python setup.

We Are Living in a Giant Casino

Short-term gratification has taken over the world. We as a society, now expect everything to be instantaneous or near-instantaneous and this attitude has made its way to nearly every corner of our lives. Here is a short but definitely not exhaustive list of such examples:

Local-first GPU Cluster with nvkind and Time Splitting

You have a brand new shiny GPU and want to start experimenting with it by running some sample experiments in Kubernetes, but how would you start that. In this short tutorial, we go over how to use nvkind, the gpu-operator to start running some basic experiemtns using your new GPU. We assume that the reader already has things such as Docker, golang, and relevant drivers/systems (nvidia-ctk,…

AI's Second Act: Predictions for the Future of AI

I want to preface that any opinions expressed here are of my own and not representative of any organization or my employers. In 2023, ChatGPT was released to the world, fundamentally changing how technology is created and how we think about businesses. What once took weeks to build now takes hours to minutes with a simple prompt and a few clicks. Tools like Cursor and Codex brought AI to our…

Distributed Inference for Fun and Profit

You ever just wonder how large models serve at scale? Or how to actually go from query to answer? Over the course of this article, we will take a look at approaches to inference and explore the tradeoffs of various approaches from a technical perspective. We assume that the reader has basic knowledge of ML concepts and how Transformers work. Additionally, all of the work here is done on a single…

A Dive into GPU Math

You ever wonder what goes on when you ask ChatGPT a question and how that is served? Or what people mean when by using a A100 to train a model and the time it takes? Or even considering the levels of abstraction between the model and the hardware? This article will aim to bring light to many of the concepts related to GPUs and the math behind them. We assume that the reader has a basic…

Tokenization and Embeddings: A Primer

Lately, all we have heard about is tokenization and embeddings and the role they play in the greater LLM and AI ecosystem. These two concepts are one of the most fundamental concepts in language modeling and remain the foundation of the technology we interact with on a daily basis. In this article, we will cover some of the basics around tokenizing and embedding sequences of texts and the nuances…

Implementing Kubernetes Security: A Hands-On Approach

In a continuation of the previous article, we explore the implementation of these different examples. Specifically, we will be covering workload separation, authentication, and other hardedning techniques. This article will have an example followed by a short explanation of what and why we should do that. We assume that the reader has a basic understanding of Kubernetes topics such as pods,…

An Intro to Kubernetes Security

Kubernetes is now widely used for managing containerized applications. As more organizations adopt it, understanding its security aspects becomes crucial. This paper examines the key security challenges in Kubernetes and suggests ways to address them. Basic Concepts of Kubernetes Security Kubernetes operates across many computers, often in different locations. This spread-out nature makes security…

Horizontal and Vertical Database Scaling

In today’s day and age, organizations have more data than ever. On a smaller scale, developers are building startups serving large amounts of data to customers. All of these use cases require both the efficient storage and transmission of data. If a backend database goes down, a service can’t be used, customers can’t reach their app, and a myriad of issues appear. To accommodate…

On Travel

Flights start blending together one after one. Washington D.C. used to seem like a world away but now is a quick afternoon drive. Early morning drives to the airport only to make it through security before the sun rises and the Dunkin’ opens. Paying an inexcusable $4.18 for a 16oz. bottle of water, only to down it moments later. Those lines have described the past month making my way around…

Life For the Past 3 Months!

It’s been 2.5 months since I graduated, and approximately 3-ish months since the last computer or mathematical related post. While alot has gone on, there will definitely be a steady catch up of content that has been written but not yet uploaded. After settling down and moving, there should be much more time to write. In lieu of that however, there has been alot of time to do cool things and…

UIUCTF'23 OSINT Writeups

A long overdue set of writeups (solutions) for UIUCTF23. Recently I had the opportunity to help run UIUCTF 2023, a student-run capture the flag (CTF) event with some awesome folk in SIGPwny. The part I worked on mainly centered around open source intelligence (OSINT) techniques. In this post, we go over some solutions to the OSINT suite as some tips and tricks to approaching these challenges in…

Adversarial AI and ChatGPT: Examples From the Wild

UPDATE: This article was covered in the Fortune Magazine article “Snap’s ‘My AI’ chatbot tells users it doesn’t know their location. It does.” on 04-21-23! A short excursion into the craze of ChatGPT In this article, we cover a basic introduction into LLMs, ChatGPT, and a real life example of Snapchat. ChatGPT, the newest craze in a series of innovations regarding artificial…

3,000 Miles of Solitude

The streets are empty with the streetlights being the only company in the morning. Wet, damp fog provide the only relief in what is projected to be another scorching day when the sun comes up. I can’t see the crows screeching into the dawn, but I can hear them singing their song. It’s July in central Illinois. I lace up my white bike cleats, marred with grease marks and mud and clip…

The Influence of Lunch Items on Crypto

*This post is pending for SIGBOVIK 0x7e7* For some background information, SIGBOVIK is a yearly scientific conference hosted by Carnegie Mellon University where participants are able to flex their lesser-known research muscles. These are the dank-memers of academia. Abstract In this paper, we introduce the discovery and exploration of the influence of common every day lunch items on cryptocurrency…

Jenkins Security Vulnerability Writeup

So, how does it begin? After getting bored one day and browsing Shodan.io, one wonders what’s the worst that you can find on there? Turns out, the answer led to a 2-day bender down different rabbit holes eventually leading to this article here and a lot of vulnerabilities discovered. So let’s get started shall we? Before we do this, I do not endorse attempting to break into company…

A Discussion on Pandas and Data Mining

At the beginning of any data analytics/data science project, the most usual case is utilizing Pandas to load the data into a object called a DataFrame and perform preprocessing tasks on it. While the Pandas library is convenient and comes with a trove of useful analytic tools, it does have some inefficiencies, many due to the nature of Python. To investigate this, let’s look at how Pandas…

Revisiting a Summer of Side Projects

This is a long overdue writeup with open source learning and Summer of Side Projects 2022 “Go do a project.” Sure, that’s what everyone talks about and describes as the optimal way to break into the tech space. Nowadays, tech seems to be all the rage, co-inflating into this large bubble of glamorous perks and lifestyles. Silicon Valley and other areas have become the new meccas for…

Inequality Fun and Their Applications

This article assumes the reader has basic knowledge of calculus and analysis. Covers topics ranging from basic Cauchy Schwarz to Lipschitz, culminating in us pulling out a mathematical rabbit out of an algebraic hat. 1. What is an Inequality? An inequality compares two values, showing if one is less than, greater than, or simply not equal to another value. Literally, it does not get more simpler…

A (Fun) Discussion on Harmonic Functions

Some introductions into harmonic functions. This post is meant to be an approchable read into these phenomena. 1. Introduction on Harmonic Functions A mass on a spring, measuring fluid flow, even the cosine wave, are all examples of where harmonic functions can be found within our world. The name itself stems from an object or force going through some type of harmonic (or oscillating) motion from…

Donuts

“$7.13 is your total sir.” These are the same words I hear every Friday morning from the woman behind the counter. Her eyes weary from the same cycle day over day; something that has become a habit over years of monotony. However, it’s not her which gives the donut shop the character it possesses; instead, it’s moreso the cacophony of hearing the ovens blaring, cash registers ringing,…

Unification in Academic Literature

Libraries store the energy that fuels the imagination. They open up windows to the world and inspire us to explore and acheive, and contribute to improving our quality of life -Sidney Sheldon Chaos is always present. Whether it’s in the weather or in crowds, things always tend to be in disarray. The same can be said for data storage and retrieval systems as well. The concept of data is odd.

A Semester of Research

All of this comes together for a Fake Detection Network, or FaDe-Net. In order to validate and benchmark our results, we performed a series of metric tests on the MIT Fake News Dataset1 and achieved a precision rate of 90.31% as well as an accuracy rate of 88.56%. Our experiments indicate the model is more effective at fake news detection. However, this model still has a significant amount of room…

Building an NLP-Powered Repository for Cyber Risk Literature

Building an NLP-Powered Repository for Cyber Risk Literature [Poster] David An, Linfeng Zheng, Zhiyu (Frank) Quan Abstract With the large and growing body of cyber risk literature, we see three major challenges faced by the actuarial research community: there is no context aware tool for finding cyber literature, no central repository of cyber risk resources, and a lack of accounting of literature…

Scraping Web Data for Academic Literature

Scraping Web Data for Academic Literature Done in collaboration with Dr. Emiliano Valdez1, University of Connecticut Goals How obvious are actual trends in academic literature and how can we easily gather academic literature metadata? The purpose of the exploratory study was to develop and analyze different methods of data retrieval from multiple sources. Given a text search within context, we…

Fake News Detection Using NLP (FaDe-Net)

FaDe-Net [Writeup] David An - AP Research Project Abstract The rapid development of social media and online news outlets has accelerated the spread of fake news across the internet. The accessibility and convenience of social media has further driven the drastic change of information consumption. As a consequence, fake news has become a significant concern because of 1) its inevitable exposure to…

(untitled)

News Archive [2-21-2024] I decided to start documenting some of the books and content I am reading into a reading list here. The intent for this is to become a centralized place for all of my various readings + thoughts. [04-21-2023] I was featured on the front page of Fortune Magazine for some interesting research into the security concerns of Snapchat's MyAI! Read the article on Fortune!…

(untitled)

Data Scientist Intern @ Ameren Innovation Center ⚡️Check back for updates on the awesome things I’m doing at Ameren…

(untitled)

Business Analyst @ D.NEXT Lab 🚚 Check back later for some cool things I did at D.NEXT!

(untitled)

Extended Readings + Reviews For a condensed list view of all my readings, please visit this link here Technical Books The DevOps Handbook, Gene Kim and Others Kim's DevOps Handbook presents itself as a solid guide to get a comprehensive understanding of the state of DevOps. The book presents itself as a wide variety of topics with real case studies baked into each chapter. While it doesn't get…

(untitled)

All Readings For extended reviews on my readings, please visit this link here. Note that the titles in bold are the ones currently being read. Technical Books The DevOps Handbook, Gene Kim and Others Building Microservices, Sam Newman Robust Python, Patrick Viafore Zero Trust Networks, Evan Gilman and Doug Barth Building Secure and Reliable Systems, Heather Adkins and Others Other Books The Design…

🌱 Digital Garden

A short collection of links that I find interesting online as well as some attached commentary or callouts with them. It is organized by year for easier navigation! 2026 2025 2025: The Year in LLMs 01-01-2026 Simon Willison looks back at the year in LLMs and the progress that has been made. Another banger piece from him Real-World Agent Examples with Gemini 3 12-21-2025 Cloudflare says service…

📚 Reading List

For extended reviews on my readings, please visit this link here. Note that the titles in bold are the ones currently being read. Technical Books The DevOps Handbook, Gene Kim and Others Building Microservices, Sam Newman Robust Python, Patrick Viafore Zero Trust Networks, Evan Gilman and Doug Barth Building Secure and Reliable Systems, Heather Adkins and Others Other Books The Design of Everyday…

Bugs

Bug Reports Got bored so here is a running of list of reports and finds in my spare time. All were found accidentally and reported to their respective security groups immediately. Organization Find Date Cloudera In progress, reported on 1-28. The fix and patch is still in progress. 1-28 AllenAI An exposed localhost port allows an actor to access production builds as an elevated user. No user…

Deploying ENS on a Private Network

This passage was taken from a GitHub repository which no longer exists and is simply on here for reference purpose We will cover the following: ​ Basic ENS architecture Setting up ENS: Deploying the registry, registrar, and resolver contracts Writing to ENS: Registering a name and storing data Reading data from ENS, both on- and off-chain ​ 1. Introduction to the ENS architecture Before going into…

Favorite Tech of 2024

2024 was the year in which we saw the capabilities of Large Langauge Models (LLMs) proliferate and mature in ways that no one has expected. We saw some historical security news and events happen as we keep learning how to keep systems secure when more of the world comes online. In here, I cover some of my favorite articles, podcasts, and books, as well as some of my (fun!) predictions for 2025.

Guide to Research Park

Why Did We Make This? We created this guide to address the growing interest in Research Park and the companies within. Recently, there has been an increased interest in all aspects of Research Park, from working there to internships, and even attending various sponsored events. However, getting involved may seem daunting at some times for students. This guide is to help anyone navigate the…

How I Made This Site

This article is part one of a two-article series. In the first part, we will focus on setting up the bare basics of how to set up a site and creating articles to view for yourself. In the second part, we focus on the basics of deploying it for the whole internet to see. Throughout the articles, we try to make it as easy as possible, so anyone without a heavy technical background can set one up…

How Many Chairs? Excursions into Fermi Questions

How many chairs are there really on campus?

PM + Software Resources

About Me David An ✍🏻 - I recently graduated with a degree in Mathematics at the University of Illinois and also currently work as a software engineer! I have past experience in creating material directed towards undergraduate recruiting and early stage careers. I am passionate about online branding, creating communities around different ideas and digital storytelling. Recruiting Resources…

PM Resources

About Us David An ✍🏻 - I recently graduated with a degree in Mathematics at the University of Illinois and also currently work as a software engineer as well! I have past experience in creating material directed towards undergraduate recruiting and early stage careers. I am passionate about online branding, creating communities around different ideas and digital storytelling. Understanding How PM…

Sitemap

Fun Stuff R-Guide For Statistics Summer of Side Projects 2022 Guide to Research Park 1.0 Blog Adversial AI and Attacking Snapchat’s My AI 3,000 Miles of Solitude The Influence of Lunch Items on Crypto Jenkins Security Vulnerability Writeup A Discussion on Pandas and Data Mining A Discussion on Pandas and Data Mining (WIP) Revisiting a Summer of Side Projects Inequality Fun and Their…

The Beauty of Polynomials

The Illusion of Progress

The first time, it was magical. You spin up a terminal, typed claude which featured a cute orange fellow dancing up welcoming you to Claude Code. Every character you type, prompt you send, and image you paste feels like progress forward. In fact, it really did feel like it. After a certain point, you reach a flow state, creating thousands of lines of code in an afternoon. Now everything feels…

The Mathematics of Strategic Decision Making

🖍️ Blog

Some excursions into different spaces. I occasionally write about math and computers. An Introduction to vLLM 02-04-2026 Fine Tuning Llama 3.2B with Unsloth and PEFT LoRA 01-12-2026 Local-first GPU Cluster with nvkind and time splitting 12-14-2025 AI’s Second Act 11-23-2025 A Dive Into GPU Math 10-15-2025 Tokenization and Embeddings, A Primer 10-01-2025 An Intro to Kubernetes Security Pt. 2…

🪷 About Me

I graduated from the University of Illinois Urbana Champaign with a B.S. in Mathematics, finishing my degree in two years (2021-2023). During that time, I organized and hosted many events catered to undergraduates in the topics of recruiting, job search, and extracurriculars. Email: davidzhongtaian(at)gmail(dot)com Linkedin: @davidzan News Features [2023] Fortune: Snap’s ‘My AI’ tells users it…