Masked language modeling is a training technique in which selected tokens in an input sequence are hidden, and the model predicts them from the remaining context.
Authority Links
Related Terms
Ready to deploy AI and LLM workflows in production?
From generative AI strategy to custom agent development and retrieval architectures, we help you scale AI responsibly.
Discuss your AI project