English

An Experimental Design Framework for Label-Efficient Supervised Finetuning of Large Language Models

Computation and Language 2024-07-09 v3 Artificial Intelligence Machine Learning

Abstract

Supervised finetuning (SFT) on instruction datasets has played a crucial role in achieving the remarkable zero-shot generalization capabilities observed in modern large language models (LLMs). However, the annotation efforts required to produce high quality responses for instructions are becoming prohibitively expensive, especially as the number of tasks spanned by instruction datasets continues to increase. Active learning is effective in identifying useful subsets of samples to annotate from an unlabeled pool, but its high computational cost remains a barrier to its widespread applicability in the context of LLMs. To mitigate the annotation cost of SFT and circumvent the computational bottlenecks of active learning, we propose using experimental design. Experimental design techniques select the most informative samples to label, and typically maximize some notion of uncertainty and/or diversity. In our work, we implement a framework that evaluates several existing and novel experimental design techniques and find that these methods consistently yield significant gains in label efficiency with little computational overhead. On generative tasks, our methods achieve the same generalization performance with only 50%50\% of annotation cost required by random sampling.

Keywords

Cite

@article{arxiv.2401.06692,
  title  = {An Experimental Design Framework for Label-Efficient Supervised Finetuning of Large Language Models},
  author = {Gantavya Bhatt and Yifang Chen and Arnav M. Das and Jifan Zhang and Sang T. Truong and Stephen Mussmann and Yinglun Zhu and Jeffrey Bilmes and Simon S. Du and Kevin Jamieson and Jordan T. Ash and Robert D. Nowak},
  journal= {arXiv preprint arXiv:2401.06692},
  year   = {2024}
}

Comments

Accepted to Findings of the Association for Computational Linguistics: ACL 2024

R2 v1 2026-06-28T14:15:26.477Z