evaluation-suite · Source-linked discovery

Dangerous Capability Evaluations

Evaluation suite spanning dangerous capabilities including autonomy, cyber and CBRN-related domains.

Open interactive record →
OriginGoogle DeepMindTopicsBio / CBRN · Cyber · Autonomy & agentsStatuscatalogued

Can support

Not independently assessed by FronteraEval yet.

Cannot support by itself

No inference beyond the upstream source should be made until the protocol is reviewed.

Original sources