# Blog

## Posts

### Memorializing My Twitter Side-Project

### Anecdotes in Language Model Coherence

### RETRO Is Blazingly Fast

### Do Infinite Pencils Exist??

### Redefining SOTA

### Why Is VSCode Typescript Linting So Damn Fast?

### My Struggle With Probability Theory

### Ducttape: Why and How

### A Software Tester's Perspective on Statistical Learning Theory

### The Variance of Yotta Savings Accounts

### Ph.D. vs. Industry Work Dynamics

### Does Pre-training Loss Linearly Predict Downstream Accuracy?

### ML Golf

### Model Compression Story - Lightning Talk

### Koan of Fish

### Personal Research Goals

### Do We Really Need Model Compression?

### All The Ways You Can Compress BERT

### The MU Puzzle Solution

### Why Use Research Workflow Management?

### Specificity - Less Wrong

### AI and Software Patents

### Improving ML Conferences

### Responsible Conduct of Research

### Koan of Duality

### Empiricism in Math and AI

### The Metaphors of Arithmetic

### Non-Operational Words

### Koan of Chores

### People Used to Be Computers

### The Importance of Not Taking Math Too Seriously

### Koan of Will

### Multiplying Complex Numbers

### Koan of Rock

### Koan of Names

### The MU Puzzle

subscribe via RSS