Language models (LMs) are pre-trained on raw text datasets to generate text sequences token-by-token. While this approach facilitates the learning of world knowledge and reasoning, it does not explicitly optimize for linguistic competence. To bridge this gap, we propose L2T, a pre-training framework integrating Language Learning Tasks alongside standard next-token prediction. Inspired by human language acquisition, L2T transforms raw text into structured input-output pairs to provide explicit linguistic stimulation. Pre-training LMs on a mixture of raw text and L2T data not only improves overall performance on linguistic competence benchmarks but accelerates its acquisition, while maintaining competitive performance on general reasoning tasks.
@article{arxiv.2601.03448,
title = {Enhancing Linguistic Competence of Language Models through Pre-training with Language Learning Tasks},
author = {Atsuki Yamaguchi and Maggie Mi and Nikolaos Aletras},
journal= {arXiv preprint arXiv:2601.03448},
year = {2026}
}