Project VeronicaUndergraduate ML → SOTA

A continuing project of NoteNextra. Bridging the gap from foundational machine learning to state-of-the-art research, one carefully ordered article at a time.

Start with Fundamentals

Jump to The Transformer Era

View on GitHub

Fundamentals & History

From OLS and SVMs to information theory — the mathematical and statistical groundwork that every modern model still rests on.

Open the basics

Deep Neural Networks

MLPs, CNNs, RNNs, generative models, and reinforcement learning — the architectures that powered the 2012–2017 deep learning era.

Read the architectures

Computer Vision

Imported from CSE559A and CSE5519 — image formation, ViTs, NeRF, 3D Gaussian Splatting, and open-vocabulary detection.

Browse vision papers

Large Language Models

Tokenization, embeddings, scaling laws, and the conceptual scaffolding behind modern LLMs.

Open the LLM track

The Transformer Era (2017→)

Chronological walk-through of the post-Transformer era — from "Attention Is All You Need" to today's reasoning and agentic models.

Walk the timeline

Why Veronica?

Our goal is to bridge the gap for undergraduates from machine-learning lessons to SOTA.
This is a continuing project of NoteNextra.

Project Veronica re-organizes scattered course notes — primarily from the WashU CSE/Math sequence — into a single, topologically ordered curriculum. Earlier articles never assume material that hasn't been introduced yet, so a curious undergraduate can read top-to-bottom and arrive at the frontier.

How the curriculum is ordered

Fundamentals & History  ──► Deep Neural Networks ──► Computer Vision
                                  │                       │
                                  ├──────────► Large Language Models
                                  │                       │
                                  └────► The Transformer Era (2017 → present)

Each section's sidebar is a topological sort of its prerequisites; the Transformer Era track is sorted chronologically, year by year.

Acknowledgments

Source material is imported and adapted from Trance-0/NoteNextra under its original license. Each imported page links back to the upstream lecture.