Back to Academy
Relevance 7/10Training ParadigmsAdvanced6 min read
Rejection Sampling
Rejection sampling keeps model outputs that pass quality criteria and discards low-quality outputs.
Why it matters for annotators
It is common in dataset curation for SFT and evaluation pipelines.
Visual mental model
Generate outputs -> filter by rubric -> keep best samples.
Examples (bad vs good)
Scenario: Real annotation scenario involving Rejection Sampling
Bad: Labeling quickly without applying project rubric.
Good: Applying rubric criteria, documenting rationale, and escalating uncertainty.
Common mistakes
- Skipping guideline details for edge cases.
- Applying inconsistent criteria across similar samples.
- Avoiding escalation even when uncertain.
Submission checklist
- Read the latest guideline update before each batch.
- Apply rubric dimensions explicitly in each decision.
- Escalate ambiguous items with concise rationale.