Intertwine
We track 3 benchmarks that Intertwine made or helped make.
0 in use 3 with public code 0 with a leaderboard here
Benchmarks
| Name | Kind | Released | Best score | Status |
|---|---|---|---|---|
| Red Team Defense Simulator Models respond to an adaptive adversary. The evaluation measures successful defense, helpfulness and false refusals. | Jailbreaks and prompt injection | Sep 2025 | 85% | Not in cards |
| Red Team Attack Simulator Models test whether a simulated assistant follows its safety policy across three conversational turns. The evaluation records policy violations and penalizes direct harm. | Jailbreaks and prompt injection | Sep 2025 | Not in cards | |
| Security Configuration Verification Agents audit Kubernetes and Terraform configurations, identify security violations and submit fixes. The grader weights detected violations by severity and validates patches. | Secure code and review | Sep 2025 | Not in cards |