It’s Not Just Size That Matters: Small Language Models Are Also Few-Shot Learners

By Timo Schick, Hinrich Schütze. 2020

1,122 citations recorded by the source corpus.

When scaled to hundreds of billions of parameters, pretrained language models such as GPT-3 (Brown et al., 2020) achieve remarkable few-shot performance. However, enormous amounts of compute are required for training and applying such big models, resulting in a large carbon footprint and making it difficult for researchers and practitioners to use them. We show that performance similar to GPT-3 can be obtained with language models that are much “greener” in that their parameter count is several orders of magnitude smaller. This is achieved by converting textual inputs into cloze questions that contain a task description, combined with gradient-based optimization; exploiting unlabeled data gives further improvements. We identify key factors required for successful natural language understanding with small language models.

Subfields: Artificial Intelligence, Large Language Models, Prompting & Reasoning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

  1. Mohit Bansal
  2. Luke S. Zettlemoyer
  3. Percy Liang
  4. Jianfeng Gao
  5. Graham Neubig
  6. Yejin Choi
  7. Noah A. Smith
  8. Hannaneh Hajishirzi

In Large Language Models

  1. Luke S. Zettlemoyer
  2. Hinrich Schütze
  3. Hannaneh Hajishirzi
  4. Yejin Choi
  5. Timo Schick
  6. Mohit Bansal
  7. Graham Neubig
  8. Zhiyuan Liu

In Prompting & Reasoning

  1. Luke S. Zettlemoyer
  2. Hannaneh Hajishirzi
  3. Timo Schick
  4. Swaroop Mishra
  5. Michael Lewis
  6. Hinrich Schütze
  7. Denny Zhou
  8. Mohit Bansal

Citations in the corpus