specific method · filed under evaluation
Cyber range exercises
Tests end-to-end cyber operations in a realistic emulated network.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
Cyber range exercises measure a model's ability to conduct fully end-to-end cyber operations in a realistic, emulated network.
usedevaluation onlyin gpt-oss-120b and gpt-oss-20bOpenAI
Filed alongside
Other methods under evaluation :: human & real-world evaluation.
Blind side-by-side expert evaluationMulti-turn open-ended external red-teamingInternal engineering task evaluationA/B testing on real tasksAutomated red-teaming loopContinuous monitoringDangerous-capability uplift assessmentExternal expert review of safety evaluation methodologyHeld-out evaluationHuman evaluationPairwise comparison for Chinese writingPre-release safety evaluationReal-world codebase testingReal-world feedback refinementReward hacking mitigation in evaluationsSame-prompt comparative DevOps testingSystematic robustness-boundary characterizationWhite-collar enterprise task evaluation