Readability $\ne$ Learnability: Rethinking the Role of Simplicity in Training Small Language Models

Open in new window