specific method · filed under evaluation
Automated red-teaming loop
Uses an attacker model to iteratively rewrite injected instructions against a defender model.
- 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.
we use an automated red-teaming loop in which an attacker model iteratively rewrites the injected instruction against a defender model
usedunclearin Agentic safety teacher for Nemotron 3 UltraNVIDIA
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 tasksContinuous monitoringCyber range exercisesDangerous-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