masked language modeling
Masked language modeling is a natural language processing technique where a specific word or token is randomly replaced with a "mask" symbol, and a model is trained to predict the original masked word based on the context of the sentence.
Requires login.
Related Concepts (1)
Similar Concepts
- computational linguistics with transformer models
- hidden markov models
- implicit feedback modeling
- language modeling
- large language model
- linguistic behavior modeling
- masking
- masking in communication
- masking in computer vision
- masking in psychology
- masking in security
- masking techniques
- role modeling
- speech recognition using transformer models
- symbolic modeling