Bad numbers in the “gzip beats BERT” paper?
The recent paper, “Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors by Jiang et al. [ link ] recently received a lot of attention on twitter.
Contact me for possible full- or part-time work. I enjoy solving difficult problems. Areas where I have spent significant time include,
Visit kenschutte.com → On Hacker News Jul 2023 – Mar 2026
Databases & Data 2AI & Machine Learning 2Software Engineering 1Life, Learning & Writing 1
The recent paper, “Low-Resource” Text Classification: A Parameter-Free Classification Method with Compressors by Jiang et al. [ link ] recently received a lot of attention on twitter.
Summary : Update to my last post analyzing the gzip+knn paper. Some of the datasets have heavily contaminated train/test tests. I show some ways to significantly speed-up the gzip method. I present more results of my "fair" reimplementation .
In 2025, I traveled by bus in South America, starting in Lima, Peru and ending in Rio de Janeiro, Brazil. Here are some details for anyone interested in doing the same.
Summary : I present a simple Python module that after imported, changes the values of two integers. The purpose is (a) show that modules can do harmful and unexpected things, (b) show an example of how to write a simple module in C, and (c) introduce a few things in CPython source code.