zero-shot

After one-shot learning.

zero-shot (not comparable)

1. (machine learning, attributive) Relating to zero-shot learning; performing without specialized training data.

   Coordinate terms: one-shot, few-shot

   zero-shot object detection ― detection of objects not in training data

   zero-shot translation ― translation between two languages without bilingual data

   • 2022, Sandra Kublik, Shubham Saboo, GPT-3, O'Reilly, →ISBN, page 20:

     Ask yourself if there is a way to get a zero-shot solution. If you think that you need external examples to prime the model for your use case, think really hard.

   • 2023 October 30, Herbold et al., “A large-scale comparison of human-written versus ChatGPT-generated essays”, in Scientific Reports, volume 13, page 4:

     No additional prompts for getting the responses were used, i.e. the data was created with a basic prompt in a zero-shot scenario.