Asked about Running a safety evaluation
Why evaluate the system rather than the model?
An answer that works
Capability is a property of the whole system. A model that fails alone may succeed with tool access, retrieval and retries, so bare-model evaluation systematically understates what is actually deployed.
Reading an answer is not being able to give one. The section that teaches this is Running a safety evaluation, 4 hours of Safety & evaluation research.
Learn this properlyAlso asked about Safety & evaluation research
How do you evaluate whether a model has a dangerous capability?Why is a negative safety result weak evidence?What is red teaming?How do you make a safety evaluation's negative result meaningful?How would you handle disclosing a vulnerability you found?Your evaluation finds a model cannot complete a concerning task. What can you conclude?How would you red team an LLM application before launch?How would you test whether a model's chain of thought is faithful?