Lunch & Learn
The team holds regular internal sessions on technical topics, tools, and operational lessons. The slides, notebooks, and occasional recordings are posted here.
2026
-
· Jim Pivarski
The Transformer
This is part of our lunch study on the foundations of LLMs. This talk explains all the parts of the transformer architecture and why each is needed.
-
· Jim Pivarski
Backpropagation
This is part of our lunch study on the foundations of LLMs. This talk explains how derivatives for optimization can be computed efficiently, starting from simple numerical and symbolic approaches and building up to backpropagation.
-
· Nick Ross
Agentic Coding Tools
A map of the agentic coding tool landscape: how the tools differ, how they are used, and an opinionated take on how the team does or does not use them.
-
· Nick Ross
Metrics for Computing
What are the core metrics that we should look at to evaluate the efficiency of the programs that we engineer? What are the tools that we use to understand the inner workings of computers, both remote and local?
2025
-
· Nick Ross
Cloud Services Best Practices
When should we pay for cloud services and when should we roll our own? How do we navigate vendor lock in?
-
· Dylan Halpern
Generating Frontend Interfaces with LLMs: Comparing Tools (As of 10/2025)
LLM-based coding tools have come a long way over the past few years, and dreams of all-in-one app builders from a chat interface feel closer than ever. But how close are we?
-
· Jim Pivarski
Adaptive Basis Functions: The Missing Link between Linear Fitting and Neural Networks
A mathematician’s path from least-squares fitting to feed-forward neural networks, with worked examples and no hand-waving analogies to brains.
-
· Jim Pivarski
Why Python is Slow, and What You Can Do About It
Why does Python often feel slower than other languages? This session digs deep into what makes Python slow, exploring how language features like dynamic typing, reference counting, and the GIL contribute to performance bottlenecks.