Present Large Language Models (LLM) self-training methods always under-sample on challenging queries, leading to inadequate learning on difficult problems which limits LLMs' ability. Therefore, this work proposes a difficulty-aware self-training (DAST) framework that focuses on improving both the quantity and quality of self-generated responses on challenging queries during self-training. DAST is specified in three components: 1) sampling-based difficulty level estimation, 2) difficulty-aware data augmentation, and 3) the self-training algorithm using SFT and DPO respectively. Experiments on mathematical tasks demonstrate the effectiveness and generalization of DAST, highlighting the critical role of difficulty-aware strategies in advancing LLM self-training.
@article{arxiv.2503.09029,
title = {DAST: Difficulty-Aware Self-Training on Large Language Models},
author = {Boyang Xue and Qi Zhu and Hongru Wang and Rui Wang and Sheng Wang and Hongling Xu and Fei Mi and Yasheng Wang and Lifeng Shang and Qun Liu and Kam-Fai Wong},
journal= {arXiv preprint arXiv:2503.09029},
year = {2025}
}