TabPFN and TabICL vs. tuned XGBoost: the model that doesn't train won 14/14
The claim behind TabPFN and TabICL is that they predict on a table without ever training on it and still beat tuned boosting.
Summary
TabPFN and TabICL are designed to predict outcomes from tabular data without traditional training, and they have outperformed optimized versions of the popular XGBoost model in every test. This article discusses how these innovative models challenge conventional training methods in machine learning.
Comments (9 of ) ⌛