LLM Reference
Concepts & capability filters

pretrained

Pretraining is the initial phase where a model learns general language representations from large unlabeled corpora via self-supervision, such as next-token prediction.

Category
Not classified
Difficulty
Not classified
Aliases
None tracked
Last reviewed
2026-07-02

Key facts

  • It is what a lab does before instruction or chat tuning; the pretraining data and scale set the ceiling on a model's knowledge and capability, which later tuning shapes but rarely exceeds.

Models Mentioning pretrained(11)