AI trainer jobs
AI trainer jobs — shape how models behave (RLHF)
Rank model responses from best to worst, critique what's wrong, and write the ideal answer. This human feedback (RLHF) is how models learn to be helpful and correct — and Perbug pays you per verified item to provide it.
RLHF ranking, critique, and reference-answer writing
Domain experts (math, code, writing, science) earn premiums
Qualification + reputation route you to higher-value projects
Remote and flexible; paid per item via PayPal
Related work
rlhf rankingresponse critiquereference answersfreeform rewritedomain expert
Frequently asked
- What is an AI trainer?
- Someone who provides the human feedback that aligns AI models — ranking responses, explaining errors, and writing better answers the model can learn from.
- What background helps?
- Clear writing plus depth in a domain (mathematics, programming, law, medicine, science) qualifies you for the best-paying RLHF and expert work.
- How much can I earn?
- It depends on task type, your qualification tier, and accuracy. Expert tasks pay more; reputation unlocks private, higher-paying programs.