Model Training & Adaptation
Supervised fine-tuning
SFT
Fine-tuning with examples of desired inputs and outputs.
Example
The model studies labeled support questions paired with the answers it should produce.
Why it matters
It helps a system reproduce a desired response pattern more consistently.
You'll hear it
“These practice answers show how we want difficult cases handled.”
Can you control it?
Set by the team
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Can the system copy mistakes in the teaching answers?
- Yes. Incorrect or inconsistent examples can teach unwanted habits.
- Does it have to start from an untrained system?
- No. It adapts a system that has already learned useful patterns.
- Can teaching one style affect other tasks?
- Yes. Adaptation can change behavior beyond the specific examples used.