Asked about Designing a capability evaluation
Your evaluation finds a model cannot complete a concerning task. What can you conclude?
An answer that works
Only that it did not under the conditions tested. The strength of the conclusion depends on elicitation effort: prompting, tools, multiple attempts, fine-tuning. Report every level tried and what was not tried, and treat the result as a lower bound on capability.
Then they push further
Why report pass at k rather than average success?
A misuser gets many attempts. A task that succeeds one time in twenty is achievable in practice, and single-attempt averages hide that.
How do you make sure the evaluation itself is not harmful?
Use proxy tasks that test necessary skills without causing the harm, run agents in isolated sandboxes with logging, and control who sees detailed results.
What gets marked up
Strong answers treat negative results as lower bounds, describe elicitation effort, and mention pass at k and sandboxing.
Reading an answer is not being able to give one. The section that teaches this is Designing a capability evaluation, 5 hours of Safety & evaluation research.
Learn this properlyAlso asked about Safety & evaluation research