concepts
- Acronym vs InitialismMay 8, 2024 ET
The difference between acronyms and initialisms, inspired by Brooklyn Nine-Nine.
- BackpropagationMay 8, 2024 ET
The error correction mechanism in neural networks.
- Neural Net: Number of Attention HeadsMay 8, 2024 ET
What attention heads are and how they affect model performance.
- Neural Net: Hidden DimensionsMay 8, 2024 ET
What hidden dimensions are in transformer architectures.
- Neural Net: Number of LayersMay 8, 2024 ET
How the number of layers affects neural network depth and performance.
- Power Law InvestingMay 8, 2024 ET
The principle that your best investment outweighs all others combined.
- RAG (Retrieval-Augmented Generation)May 8, 2024 ET
Using LLMs on new data without fine-tuning.
- Stochastic Gradient DescentMay 8, 2024 ET
The optimization algorithm used to train neural networks.
- People Avoid Annoying ThingsJan 1, 2022 ET
You can count on people doing whatever they can to avoid annoying things.
- Birthday ParadoxJan 1, 2022 ET
In a room of 70 people, there's a 99% chance two share a birthday.
- ExformationJan 1, 2022 ET
Information that is intentionally excluded from communication.
- H-IndexJan 1, 2022 ET
An author-level metric measuring productivity and citation impact.
- HeraclesJan 1, 2022 ET
Fitness goals inspired by Greek mythology.
- More Stuff, More CleaningJan 1, 2022 ET
The more stuff you have, the more stuff you have to clean.
- Morpheus is Fighting NeoJan 1, 2022 ET
A Matrix observation.
- SubTropolisJan 1, 2022 ET
The world's largest underground business complex.
- Y2038 ProblemJan 1, 2022 ET
The Unix timestamp overflow coming in 2038.