Dangerous Capability Evaluations
Google DeepMind › Safety Milestones › dangerous-capability-evaluations
Record Metadata
| Record Key | dangerous-capability-evaluations |
| Entity | Google DeepMind |
| Collection | Safety Milestones(3 records total) |
| Schema | Significant safety research publications and policy milestones. |
| YAML File | packages/kb/data/things/A4XoubikkQ.yaml |
Fields
| Name | Dangerous Capability Evaluations |
| Date | Oct 2023 |
| Type | safety-eval |
| Description | Systematic evaluations for dangerous capabilities in frontier models |
| Source | arxiv.org↗ |
| Notes | Published jointly with DeepMind safety team; covers bio, cyber, autonomy, and persuasion risks |
Other Records in Safety Milestones (2)