> ## Documentation Index > Fetch the complete documentation index at: https://proto.evodesign.org/docs/llms.txt > Use this file to discover all available pages before exploring further. # Masked Models > Protein language models for embeddings, scoring, and sampling Protein language models trained to fill in masked positions using surrounding context from both directions. They produce sequence embeddings, per-position amino acid probabilities, sampled mutations, and naturalness scores. These outputs are sequence-level priors, useful for representation, local editing, and ranking rather than structural or functional validation. * **Input:** one or more protein sequences, optionally with masked positions to fill in. * **Output:** sequence embeddings, per-position amino acid probabilities, sampled mutations, or naturalness scores.

Meta AI

Biohub