ModelRefs / Domain Adaptation — AI Glossary
Domain Adaptation — AI Glossary
Further pretraining or fine-tuning a base LLM on domain-specific text to improve performance on specialized vocabularies and tasks.
Overview
Domain adaptation involves continued pretraining on domain corpora (medical, legal, code, finance) before instruction tuning. BioMedLM, MedPaLM, and Codestral follow this pattern. It improves both factual accuracy and vocabulary coverage in the target domain without full pretraining cost.
Reference details
| Topic | training |
|---|---|
| Also known as | domain fine-tuning, domain-specific LLM |
| Last reviewed | 2026-06-24 |
Related terms
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Domain Adaptation — AI Glossary.
Frequently asked questions
What is Domain Adaptation?
Further pretraining or fine-tuning a base LLM on domain-specific text to improve performance on specialized vocabularies and tasks.
Is Domain Adaptation the same as domain fine-tuning?
Yes — domain fine-tuning, domain-specific LLM are common aliases for Domain Adaptation.
What concepts are related to Domain Adaptation?
Closely related concepts include transfer learning, continual learning, sft.