@inproceedings{xia-etal-2025-rethinking,
    title = "Rethinking Data Selection at Scale: Random Selection is Almost All You Need",
    author = "Xia, Tingyu  and
      Yu, Bowen  and
      Dang, Kai  and
      Yang, An  and
      Wu, Yuan  and
      Tian, Yuan  and
      Chang, Yi  and
      Lin, Junyang",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.146/",
    doi = "10.18653/v1/2025.findings-emnlp.146",
    pages = "2698--2711",
    ISBN = "979-8-89176-335-7",
    abstract = "Supervised fine-tuning (SFT) is crucial for aligning Large Language Models (LLMs) with human instructions. The primary goal during SFT is to select a small yet representative subset of training data from the larger pool, such that fine-tuning with this subset achieves results comparable to or even exceeding those obtained using the entire dataset. However, most existing data selection techniques are designed for small-scale data pools, which fail to meet the demands of real-world SFT scenarios. In this paper, we replicated several self-scoring methods{---}those that do not rely on external model assistance{---}on two million-scale datasets, and found that nearly all methods struggled to significantly outperform random selection when dealing with such large-scale data pools. Moreover, our comparisons suggest that, during SFT, diversity in data selection is more critical than simply focusing on high-quality data. We also analyzed the limitations of several current approaches, explaining why they perform poorly on large-scale datasets and why they are unsuitable for such contexts. Finally, we found that filtering data by token length offers a stable and efficient method for improving results. This approach, particularly when training on long-text data, proves highly beneficial for relatively weaker base models, such as Llama3. The code is available at https://github.com/xiatingyu/SFT-DataSelection-at-scale."
}