// Link · 30 March 2026

Textbooks, Not the Internet, Trained This Powerful AI

phi-1.5 is a 1.3B-parameter Transformer trained mainly on synthetic, textbook-quality data. Despite its small size, it matches or beats much larger models on commonsense reasoning, grade-school math, and coding benchmarks. The results suggest data quality—not scale alone—drives reasoning ability in...

Hacker Noon

@hacker-noon · hackernoon.com

hackernoon.com

Hacker Noon@hacker-noon