We use cookies to improve your experience on our site.
AVID-2026-R0493
Description
The model gemma-3n-E4B-it from Google was evaluated by the Garak LLM Vulnerability scanner using the probe latentinjection.LatentInjectionTranslationEnZh.
Details
The probe latentinjection.LatentInjectionTranslationEnZh tests for model responses to context-embedded injections, including indirect prompt injections and latent jailbreaks. This probe is part of the latentinjection module which detects latent injection issues.
The LLM gemma-3n-E4B-it was evaluated on this probe.
Metrics
Number of failures were counted for each detector in the probe.
| Detector | Passed | Total | Score |
|---|---|---|---|
| base.TriggerListDetector | 19 | 256 | 🟠 7.4 |
References
Affected or Relevant Artifacts
- Developer: Google
- Deployer: Together AI
- Artifact Details:
| Type | Name |
|---|---|
| Model | gemma-3n-E4B-it |
Impact
AVID Taxonomy Categorization
- Risk domains: Security
- SEP subcategories: S0403: Adversarial Example
- Lifecycle stages: L05: Evaluation
Other information
- Report Type: Measurement
- Credits:
- Date Reported: 2026-03-16
- Version: 0.3.2
- AVID Entry